{"id":217012,"date":"2026-09-02T06:16:20","date_gmt":"2026-09-02T06:16:20","guid":{"rendered":"https:\/\/10pearls.com\/uk\/blog\/\/"},"modified":"2026-09-02T06:16:23","modified_gmt":"2026-09-02T06:16:23","slug":"what-is-agentic-ai","status":"publish","type":"post","link":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/","title":{"rendered":"What Is Agentic AI?"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"217012\" class=\"elementor elementor-217012\" data-elementor-post-type=\"post\">\n\t\t\t\t<div class=\"elementor-element elementor-element-38c2b0a e-flex e-con-boxed e-con e-parent\" data-id=\"38c2b0a\" data-element_type=\"container\" data-e-type=\"container\" data-settings=\"{&quot;background_background&quot;:&quot;classic&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-96a1276 elementor-widget__width-initial indigo-h1 elementor-widget elementor-widget-heading\" data-id=\"96a1276\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h1 class=\"elementor-heading-title elementor-size-default\">What is Agentic AI and Why \nit Matters for Enterprises<\/h1>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-27410da elementor-icon-list--layout-inline elementor-list-item-link-full_width elementor-widget elementor-widget-icon-list\" data-id=\"27410da\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"icon-list.default\">\n\t\t\t\t\t\t\t<ul class=\"elementor-icon-list-items elementor-inline-items\">\n\t\t\t\t\t\t\t<li class=\"elementor-icon-list-item elementor-inline-item\">\n\t\t\t\t\t\t\t\t\t\t<span class=\"elementor-icon-list-text\">10Pearls Editorial Team<\/span>\n\t\t\t\t\t\t\t\t\t<\/li>\n\t\t\t\t\t\t\t\t<li class=\"elementor-icon-list-item elementor-inline-item\">\n\t\t\t\t\t\t\t\t\t\t\t<span class=\"elementor-icon-list-icon\">\n\t\t\t\t\t\t\t<svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-circle\" viewBox=\"0 0 512 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M256 8C119 8 8 119 8 256s111 248 248 248 248-111 248-248S393 8 256 8z\"><\/path><\/svg>\t\t\t\t\t\t<\/span>\n\t\t\t\t\t\t\t\t\t\t<span class=\"elementor-icon-list-text\">24 min read<\/span>\n\t\t\t\t\t\t\t\t\t<\/li>\n\t\t\t\t\t\t<\/ul>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-175e469 e-flex e-con-boxed e-con e-parent\" data-id=\"175e469\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t<div class=\"elementor-element elementor-element-0454a8b e-con-full e-flex e-con e-child\" data-id=\"0454a8b\" data-element_type=\"container\" data-e-type=\"container\" data-settings=\"{&quot;background_background&quot;:&quot;classic&quot;}\">\n\t\t\t\t<div class=\"elementor-element elementor-element-a38cfb2 elementor-widget elementor-widget-heading\" data-id=\"a38cfb2\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Summary<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0dcc491 elementor-widget elementor-widget-text-editor\" data-id=\"0dcc491\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>In this blog, we explore agentic AI from an enterprise lens, analyzing what it is and isn&#8217;t, how it works, use cases by business functions, differentiating characteristics, key definitions, and four dimensions of agentic readiness.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-de03dad e-con-full e-flex e-con e-child\" data-id=\"de03dad\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t<div class=\"elementor-element elementor-element-60b6362 e-con-full e-flex e-con e-child\" data-id=\"60b6362\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-da28836 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"da28836\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>The term Artificial Intelligence (AI) was coined in 1955, and it has been a discipline under computer science for decades. However, generative AI, or more specifically, Large Language Models (LLMs), is how it gained mainstream attention. From daily personal use to transforming enterprise workflows, GenAI has been an enormously transformative technology that rapidly evolved from generation to action \u2013 giving rise to agentic AI.<\/p><p>It started small, with AI assistants gaining some autonomy to perform tasks with the permission of humans. But as the underlying AI models, the wrappers and harnesses that turned them into functional systems, and the integration ecosystem that was already changing for GenAI sufficiently evolved, agentic AI gained more traction. This includes individual agents, multi-agent systems, and agentic AI orchestration that allows multiple agents to work in tandem and execute complex, multi-step tasks.<\/p><p>By 2025, 62% of organizations were at least experimenting with agentic AI, as per <a href=\"https:\/\/www.mckinsey.com\/capabilities\/quantumblack\/our-insights\/the-state-of-ai\" target=\"_blank\" rel=\"noopener nofollow\">McKinsey<\/a>, and as per <a href=\"https:\/\/www.gartner.com\/en\/articles\/hype-cycle-for-agentic-ai\" target=\"_blank\" rel=\"noopener nofollow\">Gartner&#8217;s<\/a> 2026 survey, 17% of organizations have already deployed agentic AI. In this blog, we aim to help enterprise stakeholders develop a realistic and pragmatic understanding of agentic AI, so they can make informed decisions about deploying enterprise AI agents and integrating <a href=\"\/artificial-intelligence\/agentic-ai-development-services\/\" target=\"_blank\" rel=\"noopener\">agentic AI systems<\/a> into their operations.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-6a023b1 indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"6a023b1\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"whatAgenticAI\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">What is agentic AI?<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-fcae77f section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"fcae77f\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Agentic AI refers to AI systems that can take goals, interpret them, plan how to reach these goals, execute the plan, and adjust their approach based on challenges and results of earlier execution, all without step-by-step human direction. This autonomy or \u201cagency\u201d of acting on behalf of a user or organization, tool usage, accessing internal and external information, taking decisions, and coordinating with other systems, is the defining characteristic of agentic AI.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0eeb0da indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"0eeb0da\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">The perceive, reason, act, learn loop<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1495a84 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"1495a84\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>While most agentic AI definitions don&#8217;t explicitly cover this loop, it&#8217;s there in the underlying mechanics of how most (not all) agentic AI systems operate. The four-stage loop essentially starts when the agentic AI system is given a task.<\/p><h4><strong>Perceive:<\/strong><\/h4><p><a href=\"https:\/\/www.ibm.com\/think\/topics\/ai-agent-perception#498277085\" target=\"_blank\" rel=\"noopener nofollow\">Agent perception<\/a> is mostly about how it gathers, interprets, and processes data from its environment. This ranges from its goals and enterprise knowledge it&#8217;s grounded in to external data it can access. The data can be in any form \u2013 policy documents, images, videos, and live data from sensors.<\/p><h4><strong>Reason:<\/strong><\/h4><p>After the goals are interpreted, the agents can reason and plan how to act on those goals. Many perception controls may be triggered after or during reasoning, as an agent learns what it needs to achieve its goal from internal data or external sources. It&#8217;s where decision-making happens in agentic AI systems.<\/p><h4><strong>Act:<\/strong><\/h4><p>Agentic AI systems can act on the decisions made at the reasoning stage, whether it was tool calling, generating a specific response, or taking a specific step in a workflow. Individual agents or entire agentic systems can be configured to act with full autonomy or under specific constraints, including Humans in the Loop (HITL).<\/p><h4><strong>Learn:<\/strong><\/h4><p>This is where agentic AI systems vary the most. Within a task, an agent can learn from<br \/>a failed step or an unhelpful tool response and may decide to re-run the loop differently to<br \/>get past it. Every genuinely agentic system does at least this much, and it&#8217;s the clearest difference between agentic AI and intelligent automation. Across tasks, some systems carry what they learned forward, writing outcomes to memory so later runs start better informed. Very few update the underlying model itself in production, and most enterprises deliberately keep it that way.<\/p><p>It&#8217;s important to understand that the loop isn&#8217;t a rigid flow. An agent might activate perception after reasoning or even change perception after it learns something new. Similarly, an act like tool use might actually be part of the perception, blurring the lines between the two.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-2d123bd indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"2d123bd\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">What agentic AI is not<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-3bcecb1 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"3bcecb1\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>An agentic AI is not:<\/p><ul><li><strong>Deterministic:<\/strong> Even with the same goal, configuration, and tools, an agent may take different paths for different runs, even if they are back-to-back. Each run requires a new output (or set of outputs) from the underlying LLMs, and each generation is fresh, and the model selects among possible next steps rather than returning a fixed one, so it can lead to different interpretations or a different sequence of actions. But a different path doesn\u2019t necessarily mean a different result. But it\u2019s important that enterprises don\u2019t expect the repeatability of RPA from agentic systems.<\/li><li><strong>Always multi-agent:<\/strong> An agentic AI system can just as easily be a single agent handling a complex workflow, instead of multiple agents working under an orchestration layer.<\/li><li><strong>A retrieval pipeline with a language model on top:<\/strong> Retrieval-augmented generation (RAG) grounds an answer in specific information, including enterprise knowledge, and provides the answer to the user. An agentic system decides what to retrieve, whether to retrieve again, and what to do next with what it found. Retrieval is one of its tools, not its purpose.<\/li><\/ul><p>Most agentic systems are also not fully autonomous. The goals they work against need to be realistic and grounded in the workflows and knowledge the system can actually reach, not open-ended instructions. They also need human involvement at defined points, for strategic alignment, compliance, and the contextual judgment a system working from enterprise data alone will not have.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bf3338f section-head-margin-bottom elementor-widget elementor-widget-image\" data-id=\"bf3338f\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img loading=\"lazy\" decoding=\"async\" width=\"860\" height=\"454\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body1.webp\" class=\"attachment-large size-large wp-image-217403\" alt=\"\" srcset=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body1.webp 860w, https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body1-300x158.webp 300w, https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body1-768x405.webp 768w\" sizes=\"auto, (max-width: 860px) 100vw, 860px\" loading=\"lazy\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-fbba342 indigo-h2 section-title-margin-bottom elementor-widget__width-initial elementor-widget elementor-widget-heading\" data-id=\"fbba342\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"aiTerms\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Key agentic AI terms &amp; their definitions<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bd65ffa section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"bd65ffa\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>A comprehensive understanding of agentic AI requires understanding the key concepts related to it.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-a4b70a3 mi-table section-head-margin-bottom elementor-widget elementor-widget-html\" data-id=\"a4b70a3\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t\t<div class=\"mi-table\">\r\n  <table>\r\n    <thead>\r\n      <tr>\r\n        <th>Term<\/th>\r\n        <th>Definition<\/th>\r\n      <\/tr>\r\n    <\/thead>\r\n    <tbody>\r\n      <tr>\r\n        <td>Agentic AI<\/td>\r\n        <td>An AI system that interprets a goal, plans how to reach it, executes that plan, and adjusts based on results, without step-by-step human direction.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>AI agent<\/td>\r\n        <td>A single unit that carries out a task with some autonomy, and the building block an agentic system is assembled from.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Agentic workflow<\/td>\r\n        <td>A multi-step process where the agent decides the sequence of steps rather than following a path mapped in advance.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Multi-agent system<\/td>\r\n        <td>Several agents with distinct responsibilities, coordinated toward one shared goal.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Orchestration<\/td>\r\n        <td>The layer that assigns work across agents and manages sequence, state, and handoffs between them.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Tool use, or function calling<\/td>\r\n        <td>The mechanism an agent uses to invoke an external system, API, or data source, either to retrieve information or to take an action.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Agentic RAG<\/td>\r\n        <td>Retrieval augmented generation where the agent decides what to retrieve, when to retrieve it, and whether to retrieve again.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Memory<\/td>\r\n        <td>The store that carries state and outcomes across steps and runs, which is what allows an agent to hold a goal across iterations.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Model Context Protocol (MCP)<\/td>\r\n        <td>An open standard for connecting agents to tools and data sources, so integrations are portable rather than tied to one platform.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Agent2Agent (A2A) Protocol<\/td>\r\n        <td>An open protocol for communication between agents, including agents built on different frameworks.<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Human-in-the-loop (HITL)<\/td>\r\n        <td>A design pattern that routes defined classes of action to a person for approval before the agent continues.<\/td>\r\n      <\/tr>\r\n    <\/tbody>\r\n  <\/table>\r\n<\/div>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bb86922 indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"bb86922\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"aiWorks\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">How does agentic AI work?<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-fd41f44 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"fd41f44\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Agentic AI works by combining a foundation model that supplies the reasoning, tool access that lets it act on other systems, memory that carries state across steps, and an orchestration layer that decides what runs when. The model provides the judgment. Everything built around it is what turns that judgment into completed work.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-2f9305c indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"2f9305c\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">The core building blocks<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-50fddb4 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"50fddb4\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\tA foundation model is responsible for the reasoning. It interprets the goal, decides what to do next, and revises that decision when results come back. This is why agentic AI vs. LLMs is a category confusion rather than a real comparison. The model is one component inside the agentic AI architecture, not the architecture itself.\n\nTool use, also called function calling, is the action mechanism. It gives the agent a defined set of things it is allowed to do, from querying a database to updating a record to triggering a downstream process, along with the schema for how to call each one. An agent&#8217;s capability ceiling is set by the tools it has, not by how capable the model is.\n\nMemory is responsible for maintaining the state of the agentic system. Without it, the agent has no record of what it already tried, and the loop degrades into a series of disconnected steps.\n\nThe orchestration layer decides what runs when. In a single-agent build, it is relatively thin. In a multi-agent system,  <a href=\"\/blog\/building-enterprise-ai-agent-frameworks\/\" target=\"_blank\" rel=\"noopener\">agentic AI orchestration<\/a> handles task assignment, sequencing, shared state, and handoffs between agents.\n\nMost teams assemble the individual building blocks using <a href=\"\/blog\/comparing-agentic-ai-frameworks-key-features-and-benefits\/\" target=\"_blank\" rel=\"noopener\">agentic AI frameworks<\/a>, which also supply the loop, state handling, and tool interfaces. This ensures that the engineering effort goes into the workflow rather than the plumbing.\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bad35df indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"bad35df\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Single agent vs. multi-agent systems<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-cfa2be9 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"cfa2be9\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Topology is one of the first real decisions in a build, and different topologies have different benefits and challenges.<\/p><p>A <strong>single agent<\/strong> handling one workflow is the simplest to build, observe, and debug. One reasoning thread, one set of tools, one place to look when something goes wrong. The limit is scope, as one agent can&#8217;t handle too many responsibilities without degrading.<\/p><p><strong>Multi-agent<\/strong> systems that have a defined hierarchy usually have a planner agent over specialist ones. It decomposes the goals and delegates tasks. This is good for sequential work with clear stages and good for accountability. But the planner can also be a bottleneck and a common point of failure.<\/p><p><strong>Decentralized or peer multi-agent<\/strong> systems let agents operate as equals, passing work between them without a central planner. It&#8217;s more resilient and better suited to work that cannot be mapped into stages in advance. But it&#8217;s also hard to predict and observe.<\/p><p>Going with a multi-agent design for a workflow that a single agent system can easily handle introduces additional coordination effort, increased cost, and a higher probability of failure, without improving capability.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1688df9 indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"1688df9\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Integration &amp; interoperability<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-8c785eb section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"8c785eb\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p><span class=\"TextRun SCXW30161360 BCX0\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"none\"><span class=\"NormalTextRun SCXW30161360 BCX0\">An agent that can reason but <\/span><span class=\"NormalTextRun SCXW30161360 BCX0\">can\u2019t<\/span><span class=\"NormalTextRun SCXW30161360 BCX0\"> reach your <\/span><span class=\"NormalTextRun ContextualSpellingAndGrammarErrorV2Themed SCXW30161360 BCX0\">systems<\/span><span class=\"NormalTextRun SCXW30161360 BCX0\"> is an expensive chatbot. <\/span><\/span><a href=\"\/artificial-intelligence\/ai-integration-services\/\" target=\"_blank\" rel=\"noopener\"><span class=\"TextRun Underlined SCXW30161360 BCX0\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"none\"><span class=\"NormalTextRun SCXW30161360 BCX0\">AI integration<\/span><\/span><\/a><span class=\"TextRun SCXW30161360 BCX0\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"none\"><span class=\"NormalTextRun SCXW30161360 BCX0\"> is where most of the engineering effort actually goes.<\/span><\/span><\/p><p>The baseline requirement is programmatic access to systems of record. Where an agent has to work through screens instead of APIs, it inherits every fragility that made UI-level automation brittle in the first place.<\/p><p>Above that sits a protocol layer that didn&#8217;t exist a few years ago. The Model Context Protocol (MCP) standardizes how agents connect to tools and data sources. The Agent2Agent (A2A) protocol standardizes how agents communicate with each other, including agents built on different frameworks by different vendors.<\/p><p>The core benefit here is portability. Integrations built against standardized systems are resilient against model, framework, and platform changes. In contrast, building integrations for one vendor\u2019s ecosystem and interface locks you in. Porting that for new models and frameworks can be costly and time-consuming.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-49f400e section-head-margin-bottom elementor-widget elementor-widget-image\" data-id=\"49f400e\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img loading=\"lazy\" decoding=\"async\" width=\"860\" height=\"454\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body2.webp\" class=\"attachment-large size-large wp-image-217404\" alt=\"\" srcset=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body2.webp 860w, https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body2-300x158.webp 300w, https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body2-768x405.webp 768w\" sizes=\"auto, (max-width: 860px) 100vw, 860px\" loading=\"lazy\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1b8083c indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"1b8083c\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"rpa\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Agentic AI vs. generative AI, AI agents &amp; RPA<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-12af79e section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"12af79e\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>These four categories get used interchangeably in vendor material, and the confusion is expensive. It leads enterprises to scope an architecture build as a model deployment, or to buy an agentic platform for a workflow that rules-based automation already handles well.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-df1e70f mi-table section-head-margin-bottom elementor-widget elementor-widget-html\" data-id=\"df1e70f\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t\t<div class=\"mi-table\">\r\n  <table>\r\n    <thead>\r\n      <tr>\r\n        <th><\/th>\r\n        <th>Traditional automation (RPA)<\/th>\r\n        <th>Generative AI<\/th>\r\n        <th>AI agent<\/th>\r\n        <th>Agentic AI<\/th>\r\n      <\/tr>\r\n    <\/thead>\r\n    <tbody>\r\n      <tr>\r\n        <td>What it does<\/td>\r\n        <td>Executes a path defined in advance<\/td>\r\n        <td>Produces an output on request<\/td>\r\n        <td>Completes a defined task with some autonomy<\/td>\r\n        <td>Pursues a goal across multiple steps<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Reasoning<\/td>\r\n        <td>None<\/td>\r\n        <td>Pattern-based, single turn<\/td>\r\n        <td>Bounded to its assigned task<\/td>\r\n        <td>Plans, and re-plans when results change<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Goal persistence<\/td>\r\n        <td>No<\/td>\r\n        <td>No<\/td>\r\n        <td>Limited to one task<\/td>\r\n        <td>Holds the goal across iterations<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Decides its own sequence<\/td>\r\n        <td>No<\/td>\r\n        <td>No<\/td>\r\n        <td>Partly, within its task<\/td>\r\n        <td>Yes<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Acts on external systems<\/td>\r\n        <td>Yes, scripted<\/td>\r\n        <td>No<\/td>\r\n        <td>Yes<\/td>\r\n        <td>Yes, and decides when and whether to<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Memory and state<\/td>\r\n        <td>Not applicable<\/td>\r\n        <td>None between prompts<\/td>\r\n        <td>Within a task<\/td>\r\n        <td>Across steps, and often across runs<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Behavior under variability<\/td>\r\n        <td>Breaks<\/td>\r\n        <td>Responds to what it's given<\/td>\r\n        <td>Degrades outside its task definition<\/td>\r\n        <td>Adapts, or escalates<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Repeatability<\/td>\r\n        <td>Deterministic by design<\/td>\r\n        <td>Varies by run<\/td>\r\n        <td>Varies by run<\/td>\r\n        <td>Varies by run and by path<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Human oversight model<\/td>\r\n        <td>Exception handling<\/td>\r\n        <td>Review before use<\/td>\r\n        <td>Review of the completed task<\/td>\r\n        <td>Checkpoints inside the workflow<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Typical failure mode<\/td>\r\n        <td>Halts when reality diverges<\/td>\r\n        <td>Confident wrong output<\/td>\r\n        <td>Task completed against the wrong intent<\/td>\r\n        <td>Wrong action taken at speed, or an unproductive loop<\/td>\r\n      <\/tr>\r\n    <\/tbody>\r\n  <\/table>\r\n<\/div>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-dbcef7b indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"dbcef7b\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Agentic AI vs. generative AI<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7ca7d07 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"7ca7d07\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p><a href=\"\/artificial-intelligence\/generative-ai-development-services\/\" target=\"_blank\" rel=\"noopener\">Generative AI<\/a> produces something and stops. You prompt it, it responds, and a person decides what to do with the response. The value sits in the output.<\/p><p>Agentic AI acts on that output and keeps going until the goal is met or a checkpoint blocks it. The value sits in the completed work.<\/p><p>For example, generative AI can draft the follow-up email. Give it the engagement data and it can analyze that too, then redraft the email based on what it found, and even suggest which accounts need a human touch. But each of those is a separate step that someone has to prompt, with the output of one carried to the next by hand. Agentic AI runs the sequence itself, pulling the engagement data from your CRM, deciding who gets a different message, updating the records, and flagging the accounts a person should handle.<\/p><p>This is a difference in system design more than model capability, since the same foundation model can sit underneath both. What changes is everything around it, including the tools it can call, the state it can carry, the loop it runs in, and the boundaries it works within. Agentic AI vs. traditional AI follows similar logic. Earlier systems classified, predicted, or generated. Agentic systems decide and act.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-80726ca indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"80726ca\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Agentic AI vs. AI agents<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-ca23e3a section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"ca23e3a\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>The distinction is between a component and a system.<\/p><p>An AI agent is a single unit that carries out a task with some autonomy. Agentic AI is the system-level property, covering one or more agents, the tools they can reach, shared memory, an orchestration layer, and governance around all of it.<\/p><p>A single agent can be agentic, albeit with a relatively narrow scope. For many workflows, a single well-scoped agent is enough and the right choice. But when enterprises say agentic AI, they typically mean an orchestrated system of multiple agents.<\/p><p>In practice, the terms get used interchangeably, including by vendors who know better. That&#8217;s tolerable in conversation and costly in a statement of work. Scoping a project for enterprise AI agents and scoping one for an agentic system are different exercises, with different integration requirements, governance obligations, and budgets. When both terms show up in a proposal, it&#8217;s worth asking which one is meant.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1e5506b indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"1e5506b\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Agentic AI vs. RPA &amp; traditional automation<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-fa192ba section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"fa192ba\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>RPA is about executing actions in sequences that have been mapped out in advance. It\u2019s deterministic in nature, highly auditable, and cost-effective, even at higher volumes. But it is also likely to fail as soon as unexpected exceptions come along and reality shifts from what is mapped out. A changed field or new document format can stop RPA runs.<br \/><br \/>Agentic AI decides its own execution path, which is why agentic systems can handle exceptions that would stop an RPA script.<br \/><br \/>The two aren&#8217;t competitors. RPA is still the right answer for high-volume, stable, rule-bound processes, and replacing working automation with an agent adds cost and non-determinism for no gain. The more common pattern in 2026 is agents handling the judgment and the exceptions, with deterministic automation handling the repeatable steps underneath.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7790b90 indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"7790b90\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Agentic AI vs. conversational AI &amp; chatbots<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9a04d83 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"9a04d83\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>A chatbot operates inside a conversation. It answers, it may retrieve information to answer better, and the conversation is where its work begins and ends. Agentic AI vs. <a href=\"\/ai-virtual-agents\/\" target=\"_blank\" rel=\"noopener\">conversational AI<\/a> comes down to whether anything happens outside the exchange.<\/p><p>An agentic system takes action in your systems. The conversation, where there is one, is just an interface.<\/p><p>The line has genuinely blurred, so the confusion is reasonable. Assistant products now ship agentic capabilities, which means the same interface might answer a question in one turn and complete a multi-step task in the next. The useful test isn&#8217;t what the product is called but what it&#8217;s permitted to do: whether it can only tell you something, or whether it can go and change something.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-8e140bc indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"8e140bc\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"example\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">What is an example of agentic AI?<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0ed82e0 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"0ed82e0\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Agentic AI examples are easier shown than defined, and most people have already used <br \/>at least one.<\/p><p>One of the clearest use cases is deep research. You give a system a research question, and it runs dozens of searches, goes through sources, identifies gaps, and searches again before giving you the final answer. Nobody directs the system to take all these steps. It decides on its own what to look for and how, based on the goal (research question) and what it discovers along the way (adjustment).<\/p><p>AI coding agents are the other widely used example. Given a task in a repository, they read <br \/>the codebase, plan a change, edit multiple files, run tests, read the failures, and revise. The test output is the feedback loop, which makes coding a natural fit for agentic AI in software development.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-642c1f2 indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"642c1f2\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Is ChatGPT an agentic AI?<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4d7c4e9 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"4d7c4e9\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Partially, the example above can help you understand the difference. The core chat experience is generative in nature. You prompt it and get a response, then you act. But its agent and research modes are agentic because the system takes multiple steps, uses tools like web search to fetch the latest information, and decides its own sequence. It\u2019s true for some other AI assistants as well.<\/p><p>On the enterprise side, many major platforms have developed and released agentic systems or at least added agentic layers into their platforms. These agentic systems are built around their own platforms and ecosystems. This includes Salesforce Agentforce, Microsoft\u2019s agent tooling, and AWS\u2019s Amazon Bedrock AgentCore. The platforms are frequently rebranding and merging their agentic tools, so the specific names and exact structures may change in the future.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-366766b section-head-margin-bottom elementor-widget elementor-widget-image\" data-id=\"366766b\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img loading=\"lazy\" decoding=\"async\" width=\"860\" height=\"454\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body3.webp\" class=\"attachment-large size-large wp-image-217405\" alt=\"\" srcset=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body3.webp 860w, https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body3-300x158.webp 300w, https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/Agentic-Ai-Blog-Body3-768x405.webp 768w\" sizes=\"auto, (max-width: 860px) 100vw, 860px\" loading=\"lazy\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-171c460 indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"171c460\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"matterForEnterprise\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Why does agentic AI matter for enterprises in 2026?<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-33cd326 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"33cd326\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Three things change when work moves from generation to action.<\/p><h4><strong>Scale stops requiring equivalent headcount:<\/strong><\/h4><p>Multi-step work that used to need a person to coordinate it can run without one. This is the clearest source of agentic AI ROI, and also the easiest to overstate, because the coordination usually turns out to be more of the job than anyone documented.<\/p><h4><strong>Speed improves at the hand-off points:<\/strong><\/h4><p>Most enterprise cycle time isn&#8217;t spent doing the work, but rather spent waiting between steps, for a review, a lookup, an approval, a system update. Agentic systems compress the waiting.<\/p><h4><strong>Task automation becomes outcome ownership:<\/strong><\/h4><p>This is a significant shift. When automating a task, the original process design remains intact. But when you give an agentic system a goal, you also need to define the outcome, the boundaries, and who is accountable if the system gets it wrong. This extends to process design, staffing models, and management structures, areas where <a href=\"\/blog\/understanding-agentic-ai-adoption\/\" target=\"_blank\" rel=\"noopener\">agentic AI adoption<\/a> suffers far more friction than it does from technical challenges.<\/p><p>A more pragmatic perspective is simply understanding that many agentic AI initiatives are unlikely to survive in production, with failure lying not with technology, but because the environment where they will be deployed simply isn\u2019t ready for them. Agentic AI trends in 2026 and beyond will be shaped just as much by agentic <a href=\"\/blog\/a-guide-to-ai-readiness-assessment-frameworks\/\" target=\"_blank\" rel=\"noopener\">AI readiness<\/a> as by the rapidly evolving underlying technologies.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-20d7fa3 indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"20d7fa3\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"useCases\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Enterprise agentic AI use cases by function<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bd285b5 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"bd285b5\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p><a href=\"\/blog\/exploring-the-potential-of-ai-agents-applications-across-industries\/\">Agentic AI benefits<\/a> concentrate in specific kinds of work.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-b6ddeae mi-table section-head-margin-bottom elementor-widget elementor-widget-html\" data-id=\"b6ddeae\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t\t<div class=\"mi-table\">\r\n  <table>\r\n    <thead>\r\n      <tr>\r\n        <th>Function<\/th>\r\n        <th>Where agents fit<\/th>\r\n        <th>Signal it's a fit<\/th>\r\n      <\/tr>\r\n    <\/thead>\r\n    <tbody>\r\n      <tr>\r\n        <td>Customer operations<\/td>\r\n        <td>End-to-end service resolution across systems, not just first-line response<\/td>\r\n        <td>High volume, multiple systems, variable judgment per case<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Finance and compliance<\/td>\r\n        <td>KYC screening, invoice processing, fraud review, audit trail assembly<\/td>\r\n        <td>Rule-heavy but exception-prone<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>IT operations<\/td>\r\n        <td>Log correlation, incident triage, automated remediation<\/td>\r\n        <td>Diagnosis requires assembling context from several places<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Software engineering<\/td>\r\n        <td>Code generation, test generation and maintenance, migration and refactoring, PR review<\/td>\r\n        <td>Well-specified work with verifiable output<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>Supply chain and logistics<\/td>\r\n        <td>Continuous forecasting, inventory optimization, rerouting against live conditions<\/td>\r\n        <td>Constant re-decisioning against changing data<\/td>\r\n      <\/tr>\r\n      <tr>\r\n        <td>HR and workforce<\/td>\r\n        <td>Scheduling, onboarding coordination, employee inquiries, compliance flagging<\/td>\r\n        <td>Repetitive coordination across disconnected systems<\/td>\r\n      <\/tr>\r\n    <\/tbody>\r\n  <\/table>\r\n<\/div>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-affb732 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"affb732\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>The pattern across all six: agentic AI returns the most value where workflows run across multiple steps, span multiple systems, require variable judgment, and currently depend on a person to coordinate them. Where any of those four is missing, something simpler is usually the better answer.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-8653bf9 indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"8653bf9\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"getWrong\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">What do enterprises get wrong about agentic AI?<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-3bb10f9 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"3bb10f9\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>The agentic AI challenges that derail deployments are rarely about model capability.<\/p><h4><strong>Treating it as a model upgrade instead of an architecture change:<\/strong><\/h4><p>The model is the easy part, and increasingly the commodity part. Orchestration, tool integration, memory, evaluation, and security are the work. Teams that scope an agentic project as a model selection exercise discover the actual scope after the budget is set.<\/p><h4><strong>Agent washing:<\/strong><\/h4><p>The term is Gartner&#8217;s, and it describes vendors relabeling <a href=\"\/chatbots\/\" target=\"_blank\" rel=\"noopener\"><span style=\"text-decoration: underline;\">chatbots<\/span><\/a>, assistants, and scripted workflows as agents without meaningful agentic capability. As per Gartner&#8217;s estimate, only about <a href=\"https:\/\/www.gartner.com\/en\/newsroom\/press-releases\/2025-06-25-gartner-predicts-over-40-percent-of-agentic-ai-projects-will-be-canceled-by-end-of-2027\" target=\"_blank\" rel=\"noopener nofollow\">130 out of thousands of agentic AI vendors<\/a> are real. Four questions separate the real thing from the relabel.<\/p><ul><li>Can it decide its own sequence, or does it follow a path you configured?<\/li><li>Can it recover from a failed step without human intervention?<\/li><li>What tools can it actually call, and can it call them in an order you didn&#8217;t anticipate?<\/li><li>Can you see a run where it did something you didn&#8217;t predict and the one where it stopped and asked?<\/li><\/ul><h4><strong>Deferring governance until something breaks:<\/strong><\/h4><p>An agent taking action in production systems is a non-human actor with credentials. That requires identity, scoped permissions, action logging, and instant revocation, designed before deployment and not after the first incident.<\/p><h4><strong>Ignoring data readiness:<\/strong><\/h4><p>Siloed, ungoverned, and contradictory data leads to confident wrong decisions produced at machine speed. This is one of the most common root causes of agentic failure and the least discussed, because it&#8217;s unglamorous and predates the project.<\/p><h4><strong>Over-automating before value is proven:<\/strong><\/h4><p>Another dominant failure pattern is deploying an agentic system across multiple workflows before validating whether or not it delivers consistent value across at least one. It\u2019s smart to start narrow, validate, and then expand.<\/p><h4><strong>Underestimating human-in-the-loop:<\/strong><\/h4><p>Fully autonomous AI systems aren&#8217;t production-ready for most enterprise use cases, and designing as though they are is how pilots become cancellations. HITL isn&#8217;t a limitation on autonomy. It&#8217;s what makes autonomy deployable.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-82e82e5 indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"82e82e5\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"fourDimensions\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">The four dimensions of agentic AI readiness<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-84c1269 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"84c1269\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Whether agentic AI works in an enterprise depends less on the technology than on the environment it lands in. Four dimensions determine that.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-de10279 indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"de10279\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Data foundation<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-a1ad14e section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"a1ad14e\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Agents need <a href=\"\/blog\/ai-data-readiness-roadmap-guidance\/\" target=\"_blank\" rel=\"noopener\">unified, governed, accessible data<\/a> across every system they&#8217;ll touch. They make decisions from what they can reach, so gaps and contradictions turn into wrong actions.<\/p><p>You&#8217;re ready when an agent can retrieve current data from every relevant system without a bespoke pipeline, when definitions of core entities are consistent across systems, when access controls are enforced at the data layer, and when lineage is traceable well enough to explain why a decision was made.<\/p><p>Not ready looks like three systems disagreeing about the same customer, with no authoritative answer as to which one is right.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f756a70 indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"f756a70\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Integration architecture<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f36d450 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"f36d450\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Agents need programmatic access to systems of record. Reasoning without reach <br \/>produces a demo.<br \/><br \/>You&#8217;re ready when the systems in scope expose APIs that support both reading and writing, when authentication for non-human callers is handled by design and not by shared credentials, and when adding a new tool connection is a configuration task. Standards including MCP and A2A matter here, because integration built against a standard survives a change of model, framework, or platform.<br \/><br \/>Not ready looks like the only path into a critical system being a screen a human clicks.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-7b2a164 indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"7b2a164\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Governance &amp; security for autonomous actors<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-08238aa section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"08238aa\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>An agent with credentials and the ability to act is a new class of actor, and most enterprise security models weren&#8217;t built with it in mind.<\/p><p>You can be considered ready when each agent has its own identity, minimum permissions are granted based on the workflow requirements, and every action is logged for audit. Specific critical actions must be diverted to humans before they are executed, and the system should be resilient enough that even if an agent is revoked, it doesn\u2019t take down the entire workflow.<\/p><p>This is also where the <a href=\"\/blog\/enterprise-ai-agent-security-lessons\/\" target=\"_blank\" rel=\"noopener\">threats specific to agentic systems<\/a> live, including prompt injection through content an agent reads, tool misuse, and unbounded loops that burn cost without producing anything. The OWASP work on agentic threats is a reasonable starting point for the full list.<\/p><p>Not ready looks like an agent operating on a service account shared with three integrations, with no log of what it did last week.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-77acd0c indigo-h3 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"77acd0c\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Organizational readiness<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bf6f399 section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"bf6f399\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>The dimension that gets assessed last and causes the most failures.<br \/><br \/>You&#8217;re ready when leadership has decided explicitly which decisions an agent may make and which it may not, when named individuals own the outcomes of agent-executed processes, when the teams working alongside agents know how to escalate and override, and when success is defined as a business outcome instead of a deployment.<br \/><br \/>Not ready looks like a working agentic system that nobody will approve for production, because no one is willing to own what it does.<br \/><br \/>Across all four, the same conclusion holds. Agentic AI is a strategic capability that requires foundation-level investment, and not a product you install on top of what you already have.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e17c4a7 indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"e17c4a7\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"whereStart\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Where to start<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-94dddfe section-head-margin-bottom elementor-widget elementor-widget-text-editor\" data-id=\"94dddfe\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>The four dimensions stated above can serve as a self-assessment. Run your intended first workflow against them, and the answer is usually clear. Either the constraint is technical, in which case it&#8217;s scopeable, or it&#8217;s foundational, in which case the agent isn&#8217;t the first thing <br \/>to build.<\/p><p>Most organizations find the same thing when they look honestly. The agentic layer is the visible part, and the readiness underneath it is the actual project. <span class=\"TextRun SCXW254753045 BCX0\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"none\"><span class=\"NormalTextRun SCXW254753045 BCX0\">That <\/span><span class=\"NormalTextRun SCXW254753045 BCX0\">isn\u2019t<\/span><span class=\"NormalTextRun SCXW254753045 BCX0\"> a reason to wait. <\/span><span class=\"NormalTextRun SCXW254753045 BCX0\">It\u2019s<\/span><span class=\"NormalTextRun SCXW254753045 BCX0\"> a reason to sequence the work properly, <\/span><span class=\"NormalTextRun ContextualSpellingAndGrammarErrorV2Themed SCXW254753045 BCX0\">because<\/span><span class=\"NormalTextRun SCXW254753045 BCX0\"> the data, integration, and governance investments that make agentic AI <\/span><span class=\"NormalTextRun SCXW254753045 BCX0\">viable<\/span><span class=\"NormalTextRun SCXW254753045 BCX0\"> pay off regardless of what the next architecture turns out to be. If you are scoping a first workflow and want a second read on where the constraint actually sits, <\/span><\/span><a class=\"Hyperlink SCXW254753045 BCX0\" href=\"\/\" target=\"_blank\" rel=\"noreferrer noopener\"><span class=\"TextRun Underlined SCXW254753045 BCX0\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"none\"><span class=\"NormalTextRun SCXW254753045 BCX0\" data-ccp-charstyle=\"Hyperlink\">10Pearls<\/span><\/span><\/a><span class=\"TextRun SCXW254753045 BCX0\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"none\"><span class=\"NormalTextRun SCXW254753045 BCX0\">\u2019 <\/span><\/span><a href=\"\/artificial-intelligence\/agentic-ai-development-services\/\" target=\"_blank\" rel=\"noopener\"><span class=\"TextRun Underlined SCXW254753045 BCX0\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"none\"><span class=\"NormalTextRun SCXW254753045 BCX0\">agentic AI development services<\/span><\/span><\/a><span class=\"TextRun SCXW254753045 BCX0\" lang=\"EN-US\" xml:lang=\"EN-US\" data-contrast=\"none\"><span class=\"NormalTextRun SCXW254753045 BCX0\"> cover assessment through production deployment.<\/span><\/span><span class=\"EOP Selected SCXW254753045 BCX0\" data-ccp-props=\"{&quot;201341983&quot;:0,&quot;335559740&quot;:276}\">\u00a0<\/span><\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-6e7b5fb indigo-h2 section-head-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"6e7b5fb\" data-element_type=\"widget\" data-e-type=\"widget\" id=\"faqs\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">FAQs<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e7e0471 elementor-widget-mobile__width-initial elementor-widget elementor-widget-n-accordion\" data-id=\"e7e0471\" data-element_type=\"widget\" data-e-type=\"widget\" data-settings=\"{&quot;default_state&quot;:&quot;expanded&quot;,&quot;max_items_expended&quot;:&quot;one&quot;,&quot;n_accordion_animation_duration&quot;:{&quot;unit&quot;:&quot;ms&quot;,&quot;size&quot;:400,&quot;sizes&quot;:[]}}\" data-widget_type=\"nested-accordion.default\">\n\t\t\t\t\t\t\t<div class=\"e-n-accordion\" aria-label=\"Accordion. Open links with Enter or Space, close with Escape, and navigate with Arrow Keys\">\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2430\" class=\"e-n-accordion-item\" open>\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"1\" tabindex=\"0\" aria-expanded=\"true\" aria-controls=\"e-n-accordion-item-2430\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> What is agentic AI in simple terms? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2430\" class=\"elementor-element elementor-element-5d239c7 e-con-full e-flex e-con e-child\" data-id=\"5d239c7\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-fd18c84 elementor-widget elementor-widget-text-editor\" data-id=\"fd18c84\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>In plain terms, you give an agentic system the outcome you want instead of the steps to get there, and it works out the steps itself. It interprets the goal, decides what to do, uses tools and data to act, and adjusts when something doesn&#8217;t work.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2431\" class=\"e-n-accordion-item\" >\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"2\" tabindex=\"-1\" aria-expanded=\"false\" aria-controls=\"e-n-accordion-item-2431\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> What is the difference between generative AI and agentic AI? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2431\" class=\"elementor-element elementor-element-63991a7 e-con-full e-flex e-con e-child\" data-id=\"63991a7\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-81fd9b8 elementor-widget elementor-widget-text-editor\" data-id=\"81fd9b8\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Generative AI produces an output and stops, leaving a person to act on it. Agentic AI acts on the output itself and continues until the goal is met or a checkpoint blocks it. The same foundation model can sit underneath both. What differs is the system built around it.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2432\" class=\"e-n-accordion-item\" >\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"3\" tabindex=\"-1\" aria-expanded=\"false\" aria-controls=\"e-n-accordion-item-2432\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> What is the difference between an AI agent and agentic AI? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2432\" class=\"elementor-element elementor-element-e858100 e-con-full e-flex e-con e-child\" data-id=\"e858100\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-258df7e elementor-widget elementor-widget-text-editor\" data-id=\"258df7e\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>An AI agent is a single unit that carries out a task with some autonomy. Agentic AI is the system-level property, covering one or more agents, the tools they use, shared memory, orchestration, and governance. The terms are often used interchangeably, but the distinction matters when scoping a build.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2433\" class=\"e-n-accordion-item\" >\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"4\" tabindex=\"-1\" aria-expanded=\"false\" aria-controls=\"e-n-accordion-item-2433\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> Is ChatGPT an agentic AI? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2433\" class=\"elementor-element elementor-element-4518dd9 e-con-full e-flex e-con e-child\" data-id=\"4518dd9\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-3413203 elementor-widget elementor-widget-text-editor\" data-id=\"3413203\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Partly. The base chat experience is generative, so you prompt it, it responds, and you decide what happens next. Its agent and research modes are agentic, because the system takes multiple steps, calls tools, and decides its own sequence. The mode matters more than the product name.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2434\" class=\"e-n-accordion-item\" >\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"5\" tabindex=\"-1\" aria-expanded=\"false\" aria-controls=\"e-n-accordion-item-2434\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> What is an example of agentic AI? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2434\" class=\"elementor-element elementor-element-b64aad7 e-con-full e-flex e-con e-child\" data-id=\"b64aad7\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-976e81a elementor-widget elementor-widget-text-editor\" data-id=\"976e81a\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Deep research tools that run many searches, read results, identify gaps, and search again are a common example. So are AI coding agents that plan a change, edit files, run tests, read the failures, and revise. In enterprises, agents handling end-to-end service resolution across multiple systems are typical.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2435\" class=\"e-n-accordion-item\" >\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"6\" tabindex=\"-1\" aria-expanded=\"false\" aria-controls=\"e-n-accordion-item-2435\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> How does agentic AI work? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2435\" class=\"elementor-element elementor-element-559dad5 e-con-full e-flex e-con e-child\" data-id=\"559dad5\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-9170114 elementor-widget elementor-widget-text-editor\" data-id=\"9170114\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>A foundation model supplies the reasoning, tool use gives it the ability to act on other systems, memory carries state across steps, and an orchestration layer decides what runs when. It takes a goal, plans an approach, acts, observes the result, and adjusts, repeating until the goal is met.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2436\" class=\"e-n-accordion-item\" >\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"7\" tabindex=\"-1\" aria-expanded=\"false\" aria-controls=\"e-n-accordion-item-2436\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> Is agentic AI the same as RPA? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2436\" class=\"elementor-element elementor-element-b436c0e e-con-full e-flex e-con e-child\" data-id=\"b436c0e\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-4101ee3 elementor-widget elementor-widget-text-editor\" data-id=\"4101ee3\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>No. RPA executes a path defined in advance and halts when reality diverges from it. Agentic AI decides the path and adapts when conditions change. RPA remains the better choice for high-volume, stable, rule-bound processes. The two frequently operate together in production.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2437\" class=\"e-n-accordion-item\" >\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"8\" tabindex=\"-1\" aria-expanded=\"false\" aria-controls=\"e-n-accordion-item-2437\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> What data does agentic AI need? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2437\" class=\"elementor-element elementor-element-14dd04f e-con-full e-flex e-con e-child\" data-id=\"14dd04f\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-f476f58 elementor-widget elementor-widget-text-editor\" data-id=\"f476f58\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Agents need current, governed, accessible data from every system they touch, with consistent definitions of core entities and access controls enforced at the data layer. Because agents act instead of just answering, poor data quality produces wrong actions, which makes data readiness a prerequisite.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t\t<details id=\"e-n-accordion-item-2438\" class=\"e-n-accordion-item\" >\n\t\t\t\t<summary class=\"e-n-accordion-item-title\" data-accordion-index=\"9\" tabindex=\"-1\" aria-expanded=\"false\" aria-controls=\"e-n-accordion-item-2438\" >\n\t\t\t\t\t<span class='e-n-accordion-item-title-header'><h3 class=\"e-n-accordion-item-title-text\"> Is agentic AI ready for production use? <\/h3><\/span>\n\t\t\t\t\t\t\t<span class='e-n-accordion-item-title-icon'>\n\t\t\t<span class='e-opened' ><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-minus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h384c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t\t<span class='e-closed'><svg aria-hidden=\"true\" class=\"e-font-icon-svg e-fas-plus\" viewBox=\"0 0 448 512\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\"><path d=\"M416 208H272V64c0-17.67-14.33-32-32-32h-32c-17.67 0-32 14.33-32 32v144H32c-17.67 0-32 14.33-32 32v32c0 17.67 14.33 32 32 32h144v144c0 17.67 14.33 32 32 32h32c17.67 0 32-14.33 32-32V304h144c17.67 0 32-14.33 32-32v-32c0-17.67-14.33-32-32-32z\"><\/path><\/svg><\/span>\n\t\t<\/span>\n\n\t\t\t\t\t\t<\/summary>\n\t\t\t\t<div role=\"region\" aria-labelledby=\"e-n-accordion-item-2438\" class=\"elementor-element elementor-element-55cc419 e-con-full e-flex e-con e-child\" data-id=\"55cc419\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-4a1bb90 elementor-widget elementor-widget-text-editor\" data-id=\"4a1bb90\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>For narrowly scoped workflows with strong data, integration, and governance foundations, yes. Fully autonomous systems operating across broad, loosely defined processes aren\u2019t production-ready for most enterprises. The practical approach is one workflow, clear boundaries, human checkpoints at consequential decisions, then expansion.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/details>\n\t\t\t\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-ff8df51 e-con-full e-flex e-con e-child\" data-id=\"ff8df51\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-3669af3 elementor-widget elementor-widget-heading\" data-id=\"3669af3\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">TABLE OF CONTENTS<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-b2cc744 elementor-widget elementor-widget-heading\" data-id=\"b2cc744\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#whatAgenticAI\">What is agentic AI?<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4a0f0d4 elementor-widget elementor-widget-heading\" data-id=\"4a0f0d4\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#aiTerms\">Key agentic AI terms &amp; their definitions<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bb5bf11 elementor-widget elementor-widget-heading\" data-id=\"bb5bf11\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#aiWorks\">How does agentic AI work?<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-51da978 elementor-widget elementor-widget-heading\" data-id=\"51da978\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#rpa\">Agentic AI vs. generative AI, AI agents &amp; RPA<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-10f567d elementor-widget elementor-widget-heading\" data-id=\"10f567d\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#example\">What is an example of agentic AI?<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f59257c elementor-widget elementor-widget-heading\" data-id=\"f59257c\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#matterForEnterprise\">Why does agentic AI matter for enterprises in 2026?<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f272bc1 elementor-widget elementor-widget-heading\" data-id=\"f272bc1\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#useCases\">Enterprise agentic AI use cases by function<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1d6993b elementor-widget elementor-widget-heading\" data-id=\"1d6993b\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#getWrong\">What do enterprises get wrong about agentic AI?<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1baa91a elementor-widget elementor-widget-heading\" data-id=\"1baa91a\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#fourDimensions\">The four dimensions of agentic AI readiness<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-db3dba0 elementor-widget elementor-widget-heading\" data-id=\"db3dba0\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#whereStart\">Where to start<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9af0d30 elementor-widget elementor-widget-heading\" data-id=\"9af0d30\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<p class=\"elementor-heading-title elementor-size-default\"><a href=\"#faqs\">FAQs<\/a><\/p>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-85ed69c section-padding e-flex e-con-boxed e-con e-parent\" data-id=\"85ed69c\" data-element_type=\"container\" data-e-type=\"container\" data-settings=\"{&quot;background_background&quot;:&quot;classic&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-7e44b28 indigo-h2 section-head-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"7e44b28\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\">Related blogs<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-0d44547 elementor-widget elementor-widget-insights_section_widget\" data-id=\"0d44547\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"insights_section_widget.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<style>\n\t\t\/* \u2500\u2500 Widget wrapper \u2500\u2500 *\/\n\t\t#isw-0d44547 {\n\t\t\tposition: relative;\n\t\t\tpadding: 0 40px; \/* space for arrows on left\/right *\/\n\t\t\tbox-sizing: border-box;\n\t\t}\n\t\t#isw-0d44547 .isw-outer {\n\t\t\tposition: relative;\n\t\t\toverflow: hidden;\n\t\t\tpadding: 20px 0;\n\t\t\tmargin: -20px 0;\n\t\t}\n\t\t#isw-0d44547 .isw-track {\n\t\t\tdisplay: flex;\n\t\t\tflex-direction: row;\n\t\t\tflex-wrap: nowrap;\n\t\t\tgap: 20px;\n\t\t\ttransition: transform .5s cubic-bezier(.25,.46,.45,.94);\n\t\t\twill-change: transform;\n\t\t\talign-items: stretch;\n\t\t\tvisibility: hidden;\n\t\t}\n\t\t#isw-0d44547 .isw-slide { flex-shrink: 0; box-sizing: border-box; display: flex; }\n\n\t\t\/* \u2500\u2500 Card \u2014 no shadow by default, only on hover \u2500\u2500 *\/\n\t\t#isw-0d44547 .isw-card {\n\t\t\tborder-radius: 16px; overflow: hidden; width: 100%; flex: 1;\n\t\t\tposition: relative; height: 380px; background: #fff;\n\t\t\tbox-shadow: none;\n\t\t\ttransition: box-shadow .3s ease;\n\t\t}\n\t\t#isw-0d44547 .isw-card:hover {\n\t\t\tbox-shadow: 0px 14px 36px -8px rgba(99,102,241,0.2);\n\t\t}\n\n\t\t\/* Layer 1 \u2014 Image *\/\n\t\t#isw-0d44547 .isw-img-wrap {\n\t\t\tposition: absolute; top: 0; left: 0;\n\t\t\twidth: 100%; height: 100%; z-index: 1;\n\t\t}\n\t\t#isw-0d44547 .isw-img-wrap img {\n\t\t\twidth: 100%; height: 100%; display: block; object-fit: cover;\n\t\t}\n\n\t\t\/* Layer 2 \u2014 Text bar *\/\n\t\t#isw-0d44547 .isw-text-wrap {\n\t\t\tposition: absolute; bottom: 0; left: 0;\n\t\t\twidth: 100%; height: 35%; z-index: 2;\n\t\t\tbackground: #fff; padding: 16px 24px 20px;\n\t\t\tbox-sizing: border-box; display: flex; flex-direction: column;\n\t\t\tgap: 8px; overflow: hidden;\n\t\t\ttransition: height .35s cubic-bezier(.4,0,.2,1);\n\t\t}\n\t\t#isw-0d44547 .isw-card:hover .isw-text-wrap { height: 100%; }\n\n\t\t#isw-0d44547 .isw-excerpt-wrap,\n\t\t#isw-0d44547 .isw-readmore-wrap { display: none; }\n\t\t#isw-0d44547 .isw-card:hover .isw-excerpt-wrap { display: block; }\n\t\t#isw-0d44547 .isw-card:hover .isw-readmore-wrap { display: block; margin-top: auto; }\n\n\t\t\/* \u2500\u2500 Variation 2 overrides \u2500\u2500 *\/\n\t\t#isw-0d44547.isw-v2 .isw-card {\n\t\t\theight: 530px;\n\t\t\tborder-radius: 15px;\n\t\t\tbox-shadow: none;\n\t\t}\n\t\t#isw-0d44547.isw-v2 .isw-card:hover {\n\t\t\tbox-shadow: 0px 14px 36px -8px rgba(99,102,241,0.2);\n\t\t}\n\t\t#isw-0d44547.isw-v2 .isw-img-wrap {\n\t\t\tposition: absolute; top: 0; left: 0;\n\t\t\twidth: 100%; height: 100%; z-index: 1;\n\t\t}\n\t\t#isw-0d44547.isw-v2 .isw-img-wrap img {\n\t\t\twidth: 100%; height: 100%;\n\t\t\tposition: absolute; top: 0; left: 0;\n\t\t\tobject-fit: cover;\n\t\t}\n\t\t#isw-0d44547.isw-v2 .isw-text-wrap {\n\t\t\tposition: absolute; bottom: 0; left: 0;\n\t\t\twidth: 100%; height: 120px; z-index: 2;\n\t\t\tpadding: 12px 20px;\n\t\t\tbox-sizing: border-box; display: flex;\n\t\t\tflex-direction: column; gap: 6px;\n\t\t\toverflow: hidden; justify-content: center;\n\t\t\ttransition: height .2s ease-in-out, padding .2s ease-in-out;\n\t\t}\n\t\t#isw-0d44547.isw-v2 .isw-card:hover .isw-text-wrap {\n\t\t\theight: 100% !important;\n\t\t\tpadding: 24px 20px !important;\n\t\t\tjustify-content: flex-start !important;\n\t\t\toverflow-y: auto !important;\n\t\t}\n\t\t#isw-0d44547.isw-v2 .isw-title { margin-bottom: 4px !important; margin-top: 0 !important; }\n\t\t#isw-0d44547.isw-v2 .isw-excerpt { margin: 6px 0 !important; }\n\t\t#isw-0d44547.isw-v2 .isw-readmore { text-decoration: none; display: inline-block; }\n\t\t#isw-0d44547.isw-v2 .isw-card:hover .isw-excerpt-wrap { display: block; }\n\t\t#isw-0d44547.isw-v2 .isw-card:hover .isw-readmore-wrap {\n\t\t\tdisplay: flex; justify-content: flex-end; margin-top: auto;\n\t\t}\n\n\t\t\/* \u2500\u2500 Tag \u2014 no underline, same for V1 and V2 \u2500\u2500 *\/\n\t\t#isw-0d44547 .isw-tag {\n\t\t\tfont-family: \"Figtree\", sans-serif !important;\n\t\t\tfont-size: 14px !important;\n\t\t\tfont-weight: 600 !important;\n\t\t\ttext-transform: uppercase;\n\t\t\topacity: 0.75;\n\t\t\tmargin: 0;\n\t\t\tline-height: 18px;\n\t\t\tborder-bottom: none !important;\n\t\t\tdisplay: block;\n\t\t}\n\n\t\t\/* \u2500\u2500 Title \u2500\u2500 *\/\n\t\t#isw-0d44547 .isw-title {\n\t\t\tfont-family: \"Figtree\", sans-serif !important;\n\t\t\tfont-size: 20px !important;\n\t\t\tfont-weight: 700 !important;\n\t\t\tline-height: 24px !important;\n\t\t\tcolor: #171616 !important;\n\t\t\tmargin: 0;\n\t\t\tdisplay: -webkit-box;\n\t\t\t-webkit-line-clamp: 3;\n\t\t\t-webkit-box-orient: vertical;\n\t\t\toverflow: hidden;\n\t\t}\n\t\t#isw-0d44547 .isw-title a { color: inherit !important; text-decoration: none; }\n\n\t\t\/* \u2500\u2500 Arrows \u2014 outside the slider via widget padding \u2500\u2500 *\/\n\t\t#isw-0d44547 .isw-prev,\n\t\t#isw-0d44547 .isw-next {\n\t\t\tposition: absolute;\n\t\t\ttop: calc(50% - 30px); \/* vertically centered on card area *\/\n\t\t\ttransform: translateY(-50%);\n\t\t\tbackground: none !important;\n\t\t\tborder: none !important;\n\t\t\tbox-shadow: none !important;\n\t\t\tpadding: 0;\n\t\t\tcursor: pointer;\n\t\t\twidth: 24px; height: 24px;\n\t\t\tdisplay: flex;\n\t\t\talign-items: center; justify-content: center;\n\t\t\tz-index: 10;\n\t\t\toutline: none;\n\t\t}\n\t\t#isw-0d44547 .isw-prev { left: 4px; }\n\t\t#isw-0d44547 .isw-next { right: 4px; }\n\t\t#isw-0d44547 .isw-prev::after,\n\t\t#isw-0d44547 .isw-next::after,\n\t\t#isw-0d44547 .isw-prev::before,\n\t\t#isw-0d44547 .isw-next::before { display: none !important; content: none !important; }\n\t\t@media(max-width:768px){\n\t\t\t#isw-0d44547 .isw-prev,\n\t\t\t#isw-0d44547 .isw-next { display: none !important; }\n\t\t\t#isw-0d44547 { padding: 0; }\n\t\t\t\/* V2 mobile: fixed image height, card height follows *\/\n\t\t\t#isw-0d44547.isw-v2 .isw-card { height: 500px !important; }\n\t\t\t#isw-0d44547.isw-v2 .isw-img-wrap { height: 402px !important; }\n\t\t\t#isw-0d44547.isw-v2 .isw-img-wrap img { object-fit: fill !important; height: 95% !important; }\n\t\t}\n\n\t\t\/* Dots *\/\n\t\t#isw-0d44547 .isw-dots {\n\t\t\tdisplay: flex;\n\t\t\talign-items: center;\n\t\t\tjustify-content: center;\n\t\t\tgap: 16px;\n\t\t\tmargin-top: 32px;\n\t\t\theight: 36px;\n\t\t\tbox-sizing: content-box;\n\t\t}\n\t\t#isw-0d44547 .isw-dot {\n\t\t\twidth: 10px !important;\n\t\t\theight: 10px !important;\n\t\t\tborder-radius: 50% !important;\n\t\t\tbackground: var(--e-global-color-f527530) !important;\n\t\t\topacity: 0.25;\n\t\t\tcursor: pointer;\n\t\t\tborder: none !important;\n\t\t\tpadding: 0;\n\t\t\ttransition: width .25s ease, height .25s ease, opacity .25s ease, background .25s ease, border .25s ease;\n\t\t\tflex-shrink: 0;\n\t\t\tdisplay: block;\n\t\t\tbox-sizing: border-box;\n\t\t}\n\t\t#isw-0d44547 .isw-dot.active {\n\t\t\twidth: 14px !important;\n\t\t\theight: 14px !important;\n\t\t\topacity: 1 !important;\n\t\t\tbackground: var(--e-global-color-5d532a8) !important;\n\t\t\tborder: 3px solid var(--e-global-color-f527530) !important;\n\t\t}\n\n\t\t\/* Read More *\/\n\t\t#isw-0d44547 .isw-readmore {\n\t\t\tfont-size: 14px; font-weight: 600;\n\t\t\ttext-decoration: none; color: #171616 !important;\n\t\t\tborder-bottom: 1px solid #171616;\n\t\t\tdisplay: inline-block;\n\t\t}\n\t\t<\/style><div id=\"isw-0d44547\" class=\"isw-v2\"><button class=\"isw-prev\" aria-label=\"Previous\"><svg width=\"24\" height=\"24\" viewBox=\"0 0 24 24\" fill=\"none\"><path d=\"M15.75 19.5L8.25 12L15.75 4.5\" stroke=\"#000\" stroke-width=\"1.5\" stroke-linecap=\"round\" stroke-linejoin=\"round\"\/><\/svg><\/button><button class=\"isw-next\" aria-label=\"Next\"><svg width=\"24\" height=\"24\" viewBox=\"0 0 24 24\" fill=\"none\"><path d=\"M8.25 4.5L15.75 12L8.25 19.5\" stroke=\"#000\" stroke-width=\"1.5\" stroke-linecap=\"round\" stroke-linejoin=\"round\"\/><\/svg><\/button><div class=\"isw-outer\"><div class=\"isw-track\"><div class=\"isw-slide\"><div class=\"isw-card\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/ai-vs-ml-vs-deep-learning\/\" style=\"position:absolute;top:0;left:0;width:100%;height:100%;z-index:1;display:block;text-decoration:none;\"><div class=\"isw-img-wrap\"><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/09\/AI-vs-Machine-learning-Featured-1024x1024.webp\" alt=\"AI vs ML vs Deep Learning an Enterprise Guide\" loading=\"lazy\"><\/div><\/a><div class=\"isw-text-wrap\"><p class=\"isw-tag\">AI\/ML<\/p><p class=\"isw-title\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/ai-vs-ml-vs-deep-learning\/\">AI vs ML vs Deep Learning an Enterprise Guide<\/a><\/p><div class=\"isw-excerpt-wrap\"><p class=\"isw-excerpt\">Synthetic identity fraud is the fastest growing financial crime in the US. Understanding why that is and what its life...<\/p><\/div><div class=\"isw-readmore-wrap\" style=\"text-align:left\"><a class=\"isw-readmore\" href=\"https:\/\/10pearls.com\/uk\/blog\/ai-vs-ml-vs-deep-learning\/\">Read More<\/a><\/div><\/div><\/div><\/div><div class=\"isw-slide\"><div class=\"isw-card\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/how-to-measure-ai-roi-enterprise-framework\/\" style=\"position:absolute;top:0;left:0;width:100%;height:100%;z-index:1;display:block;text-decoration:none;\"><div class=\"isw-img-wrap\"><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Measuring-AI-Investments-ROI-Featured-1024x1024.webp\" alt=\"Measuring AI Investments\u2019 ROI | Framework for Enterprise Leaders\" loading=\"lazy\"><\/div><\/a><div class=\"isw-text-wrap\"><p class=\"isw-tag\">AI\/ML<\/p><p class=\"isw-title\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/how-to-measure-ai-roi-enterprise-framework\/\">Measuring AI Investments\u2019 ROI | Framework for Enterprise Leaders<\/a><\/p><div class=\"isw-excerpt-wrap\"><p class=\"isw-excerpt\">Learn how to measure AI ROI with a practical framework covering cost savings, revenue growth, risk reduction, productivity, strategic value,...<\/p><\/div><div class=\"isw-readmore-wrap\" style=\"text-align:left\"><a class=\"isw-readmore\" href=\"https:\/\/10pearls.com\/uk\/blog\/how-to-measure-ai-roi-enterprise-framework\/\">Read More<\/a><\/div><\/div><\/div><\/div><div class=\"isw-slide\"><div class=\"isw-card\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/agentic-ai-implementation\/\" style=\"position:absolute;top:0;left:0;width:100%;height:100%;z-index:1;display:block;text-decoration:none;\"><div class=\"isw-img-wrap\"><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Agentic-AI-Implementation-Social.webp\" alt=\"Agentic AI Implementation: How to Build AI Agents\" loading=\"lazy\"><\/div><\/a><div class=\"isw-text-wrap\"><p class=\"isw-tag\">AI\/ML<\/p><p class=\"isw-title\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/agentic-ai-implementation\/\">Agentic AI Implementation: How to Build AI Agents<\/a><\/p><div class=\"isw-excerpt-wrap\"><p class=\"isw-excerpt\">Turn agentic AI from an experimental concept into a production-ready capability with guidance on architecture, development, evaluation, deployment, observability, and...<\/p><\/div><div class=\"isw-readmore-wrap\" style=\"text-align:left\"><a class=\"isw-readmore\" href=\"https:\/\/10pearls.com\/uk\/blog\/agentic-ai-implementation\/\">Read More<\/a><\/div><\/div><\/div><\/div><div class=\"isw-slide\"><div class=\"isw-card\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/\" style=\"position:absolute;top:0;left:0;width:100%;height:100%;z-index:1;display:block;text-decoration:none;\"><div class=\"isw-img-wrap\"><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Agentic-Ai-Blog-Featured-1024x1024.webp\" alt=\"What Is Agentic AI?\" loading=\"lazy\"><\/div><\/a><div class=\"isw-text-wrap\"><p class=\"isw-tag\">AI\/ML<\/p><p class=\"isw-title\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/\">What Is Agentic AI?<\/a><\/p><div class=\"isw-excerpt-wrap\"><p class=\"isw-excerpt\">Take agentic AI from promising idea to production-ready capability with a practical framework for building reliable agents, managing risk, and...<\/p><\/div><div class=\"isw-readmore-wrap\" style=\"text-align:left\"><a class=\"isw-readmore\" href=\"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/\">Read More<\/a><\/div><\/div><\/div><\/div><div class=\"isw-slide\"><div class=\"isw-card\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/ai-integration-strategy-enterprise-framework\/\" style=\"position:absolute;top:0;left:0;width:100%;height:100%;z-index:1;display:block;text-decoration:none;\"><div class=\"isw-img-wrap\"><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Building-an-AI-Integration-Strategy-Blog-Thumbnail-1024x1024.webp\" alt=\"Building an AI Integration Strategy\" loading=\"lazy\"><\/div><\/a><div class=\"isw-text-wrap\"><p class=\"isw-tag\">AI\/ML<\/p><p class=\"isw-title\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/ai-integration-strategy-enterprise-framework\/\">Building an AI Integration Strategy<\/a><\/p><div class=\"isw-excerpt-wrap\"><p class=\"isw-excerpt\">Learn a practical 9-step AI integration framework to define outcomes, overcome organizational barriers, measure ROI, and build AI solutions that...<\/p><\/div><div class=\"isw-readmore-wrap\" style=\"text-align:left\"><a class=\"isw-readmore\" href=\"https:\/\/10pearls.com\/uk\/blog\/ai-integration-strategy-enterprise-framework\/\">Read More<\/a><\/div><\/div><\/div><\/div><div class=\"isw-slide\"><div class=\"isw-card\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/ai-agent-authorization\/\" style=\"position:absolute;top:0;left:0;width:100%;height:100%;z-index:1;display:block;text-decoration:none;\"><div class=\"isw-img-wrap\"><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/AI-Agent-Authorization-Featured-1024x1024.webp\" alt=\"AI Agent Authorization: Governing Autonomous AI\" loading=\"lazy\"><\/div><\/a><div class=\"isw-text-wrap\"><p class=\"isw-tag\">AI\/ML<\/p><p class=\"isw-title\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/ai-agent-authorization\/\">AI Agent Authorization: Governing Autonomous AI<\/a><\/p><div class=\"isw-excerpt-wrap\"><p class=\"isw-excerpt\">As AI agents are granted more autonomy across enterprise systems, organizations need to define what agents can access and what...<\/p><\/div><div class=\"isw-readmore-wrap\" style=\"text-align:left\"><a class=\"isw-readmore\" href=\"https:\/\/10pearls.com\/uk\/blog\/ai-agent-authorization\/\">Read More<\/a><\/div><\/div><\/div><\/div><div class=\"isw-slide\"><div class=\"isw-card\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/agentic-ai-telecom\/\" style=\"position:absolute;top:0;left:0;width:100%;height:100%;z-index:1;display:block;text-decoration:none;\"><div class=\"isw-img-wrap\"><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/AI-in-Telecom-Featured-1024x1024.webp\" alt=\"Agentic AI in the Telecom Industry\" loading=\"lazy\"><\/div><\/a><div class=\"isw-text-wrap\"><p class=\"isw-tag\">AI\/ML<\/p><p class=\"isw-title\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/agentic-ai-telecom\/\">Agentic AI in the Telecom Industry<\/a><\/p><div class=\"isw-excerpt-wrap\"><p class=\"isw-excerpt\">The telecom industry is embracing agentic AI for multiple operational and customer-facing use cases, while navigating legacy systems, integration, and...<\/p><\/div><div class=\"isw-readmore-wrap\" style=\"text-align:left\"><a class=\"isw-readmore\" href=\"https:\/\/10pearls.com\/uk\/blog\/agentic-ai-telecom\/\">Read More<\/a><\/div><\/div><\/div><\/div><div class=\"isw-slide\"><div class=\"isw-card\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/ai-in-real-estate-industry\/\" style=\"position:absolute;top:0;left:0;width:100%;height:100%;z-index:1;display:block;text-decoration:none;\"><div class=\"isw-img-wrap\"><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/AI-in-real-estate-thumbnail-1024x1024.webp\" alt=\"How is AI being used in real estate in 2026\" loading=\"lazy\"><\/div><\/a><div class=\"isw-text-wrap\"><p class=\"isw-tag\">AI\/ML<\/p><p class=\"isw-title\"><a href=\"https:\/\/10pearls.com\/uk\/blog\/ai-in-real-estate-industry\/\">How is AI being used in real estate in 2026<\/a><\/p><div class=\"isw-excerpt-wrap\"><p class=\"isw-excerpt\">Discover how AI is transforming real estate, from asset management to contract intelligence, with practical use cases and signs that...<\/p><\/div><div class=\"isw-readmore-wrap\" style=\"text-align:left\"><a class=\"isw-readmore\" href=\"https:\/\/10pearls.com\/uk\/blog\/ai-in-real-estate-industry\/\">Read More<\/a><\/div><\/div><\/div><\/div><\/div><\/div><div class=\"isw-dots\"><\/div><\/div><script>(function(){\n\t\t\tvar id       = \"isw-0d44547\";\n\t\t\tvar spvDesk  = 3;\n\t\t\tvar spvTab   = 2;\n\t\t\tvar gap      = 20;\n\t\t\tvar total    = 8;\n\t\t\tvar autoplay = true;\n\t\t\tvar autoSpd  = 5000;\n\t\t\tvar showDots = true;\n\n\t\t\tvar wrap  = document.getElementById(id);\n\t\t\tif(!wrap) return;\n\t\t\tvar outer  = wrap.querySelector(\".isw-outer\");\n\t\t\tvar track  = wrap.querySelector(\".isw-track\");\n\t\t\tvar slides = wrap.querySelectorAll(\".isw-slide\");\n\t\t\tvar dotsWrap = wrap.querySelector(\".isw-dots\");\n\t\t\tvar btnPrev  = wrap.querySelector(\".isw-prev\");\n\t\t\tvar btnNext  = wrap.querySelector(\".isw-next\");\n\n\t\t\tvar cur      = 0;\n\t\t\tvar timer    = null;\n\t\t\tvar spv      = spvDesk;\n\n\t\t\tfunction getSpv(){\n\t\t\t\tvar w = window.innerWidth;\n\t\t\t\tif(w <= 480)  return 1;\n\t\t\t\tif(w <= 900)  return spvTab;\n\t\t\t\treturn spvDesk;\n\t\t\t}\n\n\t\t\tfunction setSlideWidths(){\n\t\t\t\tspv = getSpv();\n\t\t\t\tvar outerW = outer.offsetWidth;\n\t\t\t\tvar slideW = (outerW - gap * (spv - 1)) \/ spv;\n\t\t\t\t[].forEach.call(slides, function(s){\n\t\t\t\t\ts.style.width    = slideW + \"px\";\n\t\t\t\t\ts.style.minWidth = slideW + \"px\";\n\t\t\t\t});\n\t\t\t}\n\n\t\t\tfunction maxPage(){ return Math.max(0, total - spv); }\n\n\t\t\tfunction goTo(n, instant){\n\t\t\t\tn = Math.max(0, Math.min(n, maxPage()));\n\t\t\t\tcur = n;\n\t\t\t\tsetSlideWidths();\n\t\t\t\tvar slideW = slides[0] ? slides[0].offsetWidth : 0;\n\t\t\t\tvar offset = cur * (slideW + gap);\n\t\t\t\ttrack.style.transition = instant ? \"none\" : \"transform .5s cubic-bezier(.25,.46,.45,.94)\";\n\t\t\t\ttrack.style.transform  = \"translateX(-\" + offset + \"px)\";\n\t\t\t\tupdDots();\n\t\t\t}\n\n\t\t\tfunction updDots(){\n\t\t\t\tif(!showDots || !dotsWrap) return;\n\t\t\t\tvar pages = maxPage() + 1;\n\t\t\t\t\/\/ Rebuild dots if count changed\n\t\t\t\tif(dotsWrap.children.length !== pages){\n\t\t\t\t\tdotsWrap.innerHTML = \"\";\n\t\t\t\t\tfor(var i=0;i<pages;i++){\n\t\t\t\t\t\t(function(idx){\n\t\t\t\t\t\t\tvar b = document.createElement(\"button\");\n\t\t\t\t\t\t\tb.className = \"isw-dot\" + (idx===0?\" active\":\"\");\n\t\t\t\t\t\t\tb.setAttribute(\"aria-label\",\"Page \"+(idx+1));\n\t\t\t\t\t\t\tb.addEventListener(\"click\",function(){ stopTimer(); goTo(idx); startTimer(); });\n\t\t\t\t\t\t\tdotsWrap.appendChild(b);\n\t\t\t\t\t\t})(i);\n\t\t\t\t\t}\n\t\t\t\t}\n\t\t\t\t[].forEach.call(dotsWrap.querySelectorAll(\".isw-dot\"),function(d,i){\n\t\t\t\t\td.classList.toggle(\"active\", i===cur);\n\t\t\t\t});\n\t\t\t}\n\n\t\t\tfunction stopTimer(){ clearInterval(timer); timer=null; }\n\t\t\tfunction startTimer(){\n\t\t\t\tif(!autoplay) return; stopTimer();\n\t\t\t\ttimer = setInterval(function(){\n\t\t\t\t\tgoTo(cur >= maxPage() ? 0 : cur + 1);\n\t\t\t\t}, autoSpd);\n\t\t\t}\n\n\t\t\tif(btnPrev) btnPrev.addEventListener(\"click\",function(){ stopTimer(); goTo(cur-1); startTimer(); });\n\t\t\tif(btnNext) btnNext.addEventListener(\"click\",function(){ stopTimer(); goTo(cur+1); startTimer(); });\n\n\t\t\t\/\/ Touch swipe\n\t\t\tvar tx=0, ty=0;\n\t\t\touter.addEventListener(\"touchstart\",function(e){ tx=e.touches[0].clientX; ty=e.touches[0].clientY; stopTimer(); },{passive:true});\n\t\t\touter.addEventListener(\"touchend\",function(e){\n\t\t\t\tvar dx=e.changedTouches[0].clientX-tx, dy=e.changedTouches[0].clientY-ty;\n\t\t\t\tif(Math.abs(dx)>Math.abs(dy)&&Math.abs(dx)>40){ goTo(dx<0?cur+1:cur-1); }\n\t\t\t\tstartTimer();\n\t\t\t},{passive:true});\n\n\t\t\t\/\/ Resize\n\t\t\tvar rt;\n\t\t\twindow.addEventListener(\"resize\",function(){\n\t\t\t\tclearTimeout(rt);\n\t\t\t\trt=setTimeout(function(){ goTo(cur,true); },200);\n\t\t\t});\n\n\t\t\t\/\/ Init \u2014 use requestAnimationFrame so layout is painted and offsetWidth is correct\n\t\t\tfunction doInit(){\n\t\t\t\tsetSlideWidths();\n\t\t\t\tupdDots();\n\t\t\t\tgoTo(0, true);\n\t\t\t\ttrack.style.visibility = \"visible\"; \/\/ show after widths are set\n\t\t\t\tstartTimer();\n\t\t\t}\n\t\t\tif(document.readyState === \"loading\"){\n\t\t\t\tdocument.addEventListener(\"DOMContentLoaded\", function(){ requestAnimationFrame(doInit); });\n\t\t\t} else {\n\t\t\t\trequestAnimationFrame(doInit);\n\t\t\t}\n\t\t})();<\/script>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-94e75c2 section-padding gradient-lefttoright-service e-flex e-con-boxed e-con e-parent\" data-id=\"94e75c2\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t<div class=\"elementor-element elementor-element-ea61807 e-con-full e-flex e-con e-child\" data-id=\"ea61807\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t<div class=\"elementor-element elementor-element-8304c2a indigo-h2 section-title-margin-bottom elementor-widget elementor-widget-heading\" data-id=\"8304c2a\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h2 class=\"elementor-heading-title elementor-size-default\"><span>Get in touch<\/span> with us<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-6374d59 elementor-widget elementor-widget-text-editor\" data-id=\"6374d59\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t\t\t\t\t\t<p>Global digital transformation and product engineering partner.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-939461a e-con-full e-flex e-con e-child\" data-id=\"939461a\" data-element_type=\"container\" data-e-type=\"container\" data-settings=\"{&quot;background_background&quot;:&quot;classic&quot;}\">\n\t\t\t\t<div class=\"elementor-element elementor-element-eb8ed13 indigo-h3 elementor-widget elementor-widget-heading\" data-id=\"eb8ed13\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Contact Information<\/h3>\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-71d0c8c elementor-widget elementor-widget-wpforms\" data-id=\"71d0c8c\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"wpforms.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<style id=\"wpforms-css-vars-elementor-widget-71d0c8c\">\n\t\t\t\t.elementor-widget-wpforms.elementor-element-71d0c8c {\n\t\t\t\t--wpforms-background-url: url( https:\/\/10pearls.com\/uk\/wp-content\/plugins\/elementor\/assets\/images\/placeholder.png );\n\t\t\t}\n\t\t\t<\/style><div class=\"wpforms-container wpforms-container-full wpforms-render-modern\" id=\"wpforms-215949\"><form id=\"wpforms-form-215949\" class=\"wpforms-validate wpforms-form wpforms-ajax-form\" data-formid=\"215949\" method=\"post\" enctype=\"multipart\/form-data\" action=\"\/uk\/wp-json\/wp\/v2\/posts\/217012\" data-token=\"11251715b6a5be9f4bf439ab5667b0d6\" data-token-time=\"1788360171\"><noscript class=\"wpforms-error-noscript\">Please enable JavaScript in your browser to complete this form.<\/noscript><div id=\"wpforms-error-noscript\" style=\"display: none;\">Please enable JavaScript in your browser to complete this form.<\/div><div class=\"wpforms-field-container\">\t\t<div id=\"wpforms-215949-field_1-container\"\n\t\t\tclass=\"wpforms-field wpforms-field-text\"\n\t\t\tdata-field-type=\"text\"\n\t\t\tdata-field-id=\"1\"\n\t\t\t>\n\t\t\t<label class=\"wpforms-field-label\" for=\"wpforms-215949-field_1\" >Email<\/label>\n\t\t\t<input type=\"text\" id=\"wpforms-215949-field_1\" class=\"wpforms-field-medium\" name=\"wpforms[fields][1]\" >\n\t\t<\/div>\n\t\t<div id=\"wpforms-215949-field_2-container\" class=\"wpforms-field wpforms-field-email\" data-field-type=\"email\" data-field-id=\"2\"><label class=\"wpforms-field-label\" for=\"wpforms-215949-field_2\">Email <span class=\"wpforms-required-label\" aria-hidden=\"true\">*<\/span><\/label><input type=\"email\" id=\"wpforms-215949-field_2\" class=\"wpforms-field-large wpforms-field-required\" name=\"wpforms[fields][2]\" placeholder=\"Enter you email\" spellcheck=\"false\" aria-errormessage=\"wpforms-215949-field_2-error\" required><\/div><script>\n\t\t\t\t( function() {\n\t\t\t\t\tconst style = document.createElement( 'style' );\n\t\t\t\t\tstyle.appendChild( document.createTextNode( '#wpforms-215949-field_1-container { position: absolute !important; overflow: hidden !important; display: inline !important; height: 1px !important; width: 1px !important; z-index: -1000 !important; padding: 0 !important; } #wpforms-215949-field_1-container input { visibility: hidden; } #wpforms-conversational-form-page #wpforms-215949-field_1-container label { counter-increment: none; }' ) );\n\t\t\t\t\tdocument.head.appendChild( style );\n\t\t\t\t\tdocument.currentScript?.remove();\n\t\t\t\t} )();\n\t\t\t<\/script><\/div><!-- .wpforms-field-container --><div class=\"wpforms-recaptcha-container wpforms-is-hcaptcha\" ><div class=\"h-captcha\" data-sitekey=\"b192f310-e2a0-450e-bda4-15f8857a8fbc\"><\/div><input type=\"text\" name=\"g-recaptcha-hidden\" class=\"wpforms-recaptcha-hidden\" style=\"position:absolute!important;clip:rect(0,0,0,0)!important;height:1px!important;width:1px!important;border:0!important;overflow:hidden!important;padding:0!important;margin:0!important;\" data-rule-hcaptcha=\"1\"><\/div><div class=\"wpforms-submit-container\" ><input type=\"hidden\" name=\"wpforms[id]\" value=\"215949\"><input type=\"hidden\" name=\"page_title\" value=\"\"><input type=\"hidden\" name=\"page_url\" value=\"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/posts\/217012\"><input type=\"hidden\" name=\"url_referer\" value=\"\"><button type=\"submit\" name=\"wpforms[submit]\" id=\"wpforms-submit-215949\" class=\"wpforms-submit indigo-btn-self indigo-btn-self--gradient\" data-alt-text=\"Sending...\" data-submit-text=\"Submit\" aria-live=\"assertive\" value=\"wpforms-submit\">Submit<\/button><img decoding=\"async\" src=\"https:\/\/10pearls.com\/uk\/wp-content\/plugins\/wpforms\/assets\/images\/submit-spin.svg\" class=\"wpforms-submit-spinner\" style=\"display: none;\" width=\"26\" height=\"26\" alt=\"Loading\"><\/div><\/form><\/div>  <!-- .wpforms-container -->\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-4375fc9 e-flex e-con-boxed e-con e-parent\" data-id=\"4375fc9\" data-element_type=\"container\" data-e-type=\"container\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>Take agentic AI from promising idea to production-ready capability with a practical framework for building reliable agents, managing risk, and delivering measurable business value.<\/p>\n","protected":false},"author":85,"featured_media":217017,"comment_status":"closed","ping_status":"open","sticky":false,"template":"elementor_theme","format":"standard","meta":{"inline_featured_image":false,"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[49],"tags":[4585],"class_list":["post-217012","post","type-post","status-publish","format-standard","has-post-thumbnail","category-blog","tag-ai-and-ml"],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v25.5 (Yoast SEO v28.4) - https:\/\/yoast.com\/product\/yoast-seo-premium-wordpress\/ -->\n<title>What Is Agentic AI?<\/title>\n<meta name=\"description\" content=\"Agentic AI plans and acts toward a goal without step-by-step prompts. How it differs from generative AI and AI agents, and what enterprises need to get right.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/\" \/>\n<meta property=\"og:locale\" content=\"en_GB\" \/>\n<meta property=\"og:type\" content=\"website\" \/>\n<meta property=\"og:title\" content=\"What Is Agentic AI?\" \/>\n<meta property=\"og:description\" content=\"Agentic AI plans and acts toward a goal without step-by-step prompts. How it differs from generative AI and AI agents, and what enterprises need to get right.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/\" \/>\n<meta property=\"og:site_name\" content=\"10Pearls UK\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/10Pearls\/\" \/>\n<meta property=\"article:published_time\" content=\"2026-09-02T06:16:20+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-09-02T06:16:23+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Agentic-Ai-Blog-Social.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"640\" \/>\n\t<meta property=\"og:image:height\" content=\"359\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"author\" content=\"Bilal Khalid\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:title\" content=\"What Is Agentic AI?\" \/>\n<meta name=\"twitter:description\" content=\"Agentic AI plans and acts toward a goal without step-by-step prompts. How it differs from generative AI and AI agents, and what enterprises need to get right.\" \/>\n<meta name=\"twitter:creator\" content=\"@TenPearls\" \/>\n<meta name=\"twitter:site\" content=\"@TenPearls\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Bilal Khalid\" \/>\n\t<meta name=\"twitter:label2\" content=\"Estimated reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"27 minutes\" \/>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"What Is Agentic AI?","description":"Agentic AI plans and acts toward a goal without step-by-step prompts. How it differs from generative AI and AI agents, and what enterprises need to get right.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/","og_locale":"en_GB","og_type":"website","og_title":"What Is Agentic AI?","og_description":"Agentic AI plans and acts toward a goal without step-by-step prompts. How it differs from generative AI and AI agents, and what enterprises need to get right.","og_url":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/","og_site_name":"10Pearls UK","article_publisher":"https:\/\/www.facebook.com\/10Pearls\/","article_published_time":"2026-09-02T06:16:20+00:00","article_modified_time":"2026-09-02T06:16:23+00:00","og_image":[{"width":640,"height":359,"url":"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Agentic-Ai-Blog-Social.webp","type":"image\/webp"}],"author":"Bilal Khalid","twitter_card":"summary_large_image","twitter_title":"What Is Agentic AI?","twitter_description":"Agentic AI plans and acts toward a goal without step-by-step prompts. How it differs from generative AI and AI agents, and what enterprises need to get right.","twitter_creator":"@TenPearls","twitter_site":"@TenPearls","twitter_misc":{"Written by":"Bilal Khalid","Estimated reading time":"27 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/#article","isPartOf":{"@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/"},"author":{"name":"Bilal Khalid","@id":"https:\/\/10pearls.com\/uk\/#\/schema\/person\/30a3e41d42f83257c851d0b716a592ed"},"headline":"What Is Agentic AI?","datePublished":"2026-09-02T06:16:20+00:00","dateModified":"2026-09-02T06:16:23+00:00","mainEntityOfPage":{"@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/"},"wordCount":5646,"publisher":{"@id":"https:\/\/10pearls.com\/uk\/#organization"},"image":{"@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/#primaryimage"},"thumbnailUrl":"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Agentic-Ai-Blog-Featured.webp","keywords":["AI\/ML"],"articleSection":["Blog"],"inLanguage":"en-GB"},{"@type":"WebPage","@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/","url":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/","name":"What Is Agentic AI?","isPartOf":{"@id":"https:\/\/10pearls.com\/uk\/#website"},"primaryImageOfPage":{"@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/#primaryimage"},"image":{"@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/#primaryimage"},"thumbnailUrl":"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Agentic-Ai-Blog-Featured.webp","datePublished":"2026-09-02T06:16:20+00:00","dateModified":"2026-09-02T06:16:23+00:00","description":"Agentic AI plans and acts toward a goal without step-by-step prompts. How it differs from generative AI and AI agents, and what enterprises need to get right.","breadcrumb":{"@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/#breadcrumb"},"inLanguage":"en-GB","potentialAction":[{"@type":"ReadAction","target":["https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/"]}]},{"@type":"ImageObject","inLanguage":"en-GB","@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/#primaryimage","url":"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Agentic-Ai-Blog-Featured.webp","contentUrl":"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2026\/08\/Agentic-Ai-Blog-Featured.webp","width":4096,"height":4096,"caption":"Agentic Ai Blog Featured"},{"@type":"BreadcrumbList","@id":"https:\/\/10pearls.com\/uk\/blog\/what-is-agentic-ai\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/10pearls.com\/uk\/"},{"@type":"ListItem","position":2,"name":"What Is Agentic AI?"}]},{"@type":"WebSite","@id":"https:\/\/10pearls.com\/uk\/#website","url":"https:\/\/10pearls.com\/uk\/","name":"10Pearls UK","description":"AI-Powered Software Development Company","publisher":{"@id":"https:\/\/10pearls.com\/uk\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/10pearls.com\/uk\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-GB"},{"@type":"Organization","@id":"https:\/\/10pearls.com\/uk\/#organization","name":"10Pearls","alternateName":"Software development company","url":"https:\/\/10pearls.com\/uk\/","logo":{"@type":"ImageObject","inLanguage":"en-GB","@id":"https:\/\/10pearls.com\/uk\/#\/schema\/logo\/image\/","url":"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2025\/07\/10Pearls-Black.png","contentUrl":"https:\/\/10pearls.com\/uk\/wp-content\/uploads\/2025\/07\/10Pearls-Black.png","width":282,"height":107,"caption":"10Pearls"},"image":{"@id":"https:\/\/10pearls.com\/uk\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/10Pearls\/","https:\/\/x.com\/TenPearls","https:\/\/www.linkedin.com\/company\/10pearls\/"],"description":"10Pearls is a leading AI-powered software development company. We build intelligent digital products for healthcare, finance, retail, and more.","email":"info@10pearls.com","telephone":"+44 7957 620104","legalName":"10Pearls"},{"@type":"Person","@id":"https:\/\/10pearls.com\/uk\/#\/schema\/person\/30a3e41d42f83257c851d0b716a592ed","name":"Bilal Khalid","image":{"@type":"ImageObject","inLanguage":"en-GB","@id":"https:\/\/secure.gravatar.com\/avatar\/bf7fde25f986fab31d153cf3544480147588e5e29446b5515b4fe18999d74056?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/bf7fde25f986fab31d153cf3544480147588e5e29446b5515b4fe18999d74056?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/bf7fde25f986fab31d153cf3544480147588e5e29446b5515b4fe18999d74056?s=96&d=mm&r=g","caption":"Bilal Khalid"}}]}},"_links":{"self":[{"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/posts\/217012","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/users\/85"}],"replies":[{"embeddable":true,"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/comments?post=217012"}],"version-history":[{"count":8,"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/posts\/217012\/revisions"}],"predecessor-version":[{"id":217408,"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/posts\/217012\/revisions\/217408"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/media\/217017"}],"wp:attachment":[{"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/media?parent=217012"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/categories?post=217012"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/10pearls.com\/uk\/wp-json\/wp\/v2\/tags?post=217012"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}