<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Jayan Chandran on Medium]]></title>
        <description><![CDATA[Stories by Jayan Chandran on Medium]]></description>
        <link>https://medium.com/@jayanchandran01?source=rss-212bf3e78c71------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*3t54nnA6clxQZAE2rx-D2Q.jpeg</url>
            <title>Stories by Jayan Chandran on Medium</title>
            <link>https://medium.com/@jayanchandran01?source=rss-212bf3e78c71------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 14:51:47 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@jayanchandran01/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Designing for Agents: What Designers Need to Know About Agent Experience (AX)]]></title>
            <link>https://medium.com/@jayanchandran01/designing-for-agents-what-designers-need-to-know-about-agent-experience-ax-9bebca5bfa04?source=rss-212bf3e78c71------2</link>
            <guid isPermaLink="false">https://medium.com/p/9bebca5bfa04</guid>
            <category><![CDATA[agent-experience]]></category>
            <category><![CDATA[agentic-ai]]></category>
            <category><![CDATA[ax]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[design-thinking]]></category>
            <dc:creator><![CDATA[Jayan Chandran]]></dc:creator>
            <pubDate>Mon, 12 May 2025 08:22:00 GMT</pubDate>
            <atom:updated>2025-05-12T08:22:00.172Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_v9K8AytK7sD4iYP5q-8LA.png" /></figure><p>For years, designers have focused on crafting great <strong>User Experiences (UX)</strong> — making digital products that are easy and intuitive for humans. But now, there’s a new kind of user in the mix: <strong>AI agents</strong>. With them comes a new design discipline: <strong>Agent Experience (AX)</strong>.</p><blockquote><em>“We must design for the way people behave, not for how we would wish them to behave.” — </em><strong><em>Donald A. Norman, </em></strong><a href="https://www.goodreads.com/work/quotes/13144201"><strong><em>Living with Complexity</em></strong></a></blockquote><p>So what exactly is <strong>AX</strong>, and what should designers be thinking about?</p><h3>What is AX (Agent Experience)?</h3><p>Agent Experience is about designing your digital product or service so that AI-powered agents — not just people — can use it effectively.</p><p>These agents are not just simple chatbots or rule-based scripts. They’re more like intelligent digital assistants powered by advanced AI, such as Large Language Models (LLMs). Users interact with them using natural language, and then these agents go off to complete tasks on the user’s behalf.</p><p>Instead of the user tapping buttons or filling out forms, the agent does it for them. The agent acts as a delegate — an extension of the user — interacting with your product or service behind the scenes.</p><h3>Real-Life Examples of Agents in Action</h3><p>Agents are already being used in a wide range of everyday tasks. Here are a few examples:</p><ul><li>A designer asks an agent to apply a new color scheme to a design component or run a quick usability test.</li><li>A developer uses an AI assistant to debug code or set up a server environment.</li><li>A customer support agent asks an internal bot to find relevant documentation to answer a ticket.</li><li>A parent asks an AI assistant to plan a night out, book a babysitter, and arrange transportation.</li></ul><p>These are not future scenarios — they’re happening now. And more users are beginning to rely on agents to get things done.</p><h3>Why Designers Should Care About AX</h3><p>As agents become more common, designers need to <strong>think beyond just human interfaces</strong>. If you’re only designing for humans, you’re missing a growing segment of how users will actually interact with your service.</p><p>Good AX means that agents can:</p><ul><li>Understand your product’s structure and content</li><li>Access information easily</li><li>Complete tasks on behalf of the user smoothly</li></ul><p>It’s not about replacing UX — it’s about extending it to a new type of user.</p><h3>Key Design Principles for a Great Agent Experience</h3><p>Designing for AX requires a shift in mindset. Below are some important considerations for designers to keep in mind.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BSmlfObbQvZlxVd9ROCqYQ.png" /></figure><h3>1. Design with Discoverability in Mind — for Machines</h3><p>Agents need to “see” your service in a structured way. They don’t visually scan pages like humans. They parse information, often through APIs, structured data, or context-aware interfaces.</p><p><strong><em>Tips:</em></strong></p><ul><li>Use clean, semantic HTML and structured data.</li><li>Design consistent UI patterns that are easy to interpret.</li><li>Make key content and actions easy to identify through code and markup.</li></ul><h3>2. Create Machine-Readable Interfaces and Documentation</h3><p>Unlike humans, agents don’t read help pages or hover over tooltips. They need <strong>structured, well-documented ways to access features and understand workflows.</strong></p><p><strong><em>Tips:</em></strong></p><ul><li>Provide accessible APIs and clear API documentation.</li><li>Use or prepare for emerging standards like:</li><li><strong>LLMs.txt</strong> — lets you describe your service for AI crawlers</li><li><strong>Model Context Protocol (MCP)</strong> — helps set context for how agents interact</li><li><strong>Arazzo</strong> — defines agent interaction patterns</li></ul><h3>3. Think in Tasks, Not Just Screens</h3><p>Agents work best when they understand <strong>what to do</strong> — not just where to click.</p><p><strong><em>Tips:</em></strong></p><ul><li>Break your product down into clear, goal-based tasks.</li><li>Make those tasks <strong>logical and sequential</strong> for automation.</li><li>Ensure the agent has access to all required data and steps without human hand-holding.</li></ul><h3>4. Shift from Control to Influence</h3><p>You can’t dictate every move an agent makes like you can with a user interface. Agents interpret your product based on <strong>context and signals</strong>.</p><p><strong><em>Tips:</em></strong></p><ul><li>Design experiences that provide clear <strong>intent</strong>, not rigid flows.</li><li>Offer strong guidance (through metadata, labels, or affordances) rather than tight controls.</li><li>Help agents “understand” what’s important and what actions are available.</li></ul><h3>5. Support a Range of Agent Interactions</h3><p>Agents will differ depending on the tools or services users choose. One agent might interact via your API, another might rely on scraping your website. <strong>Design for diversity.</strong></p><p><strong><em>Tips:</em></strong></p><ul><li>Be flexible in how your product can be accessed.</li><li>Ensure fallbacks exist if a structured path isn’t available.</li><li>Support graceful degradation — make sure agents can at least gather or summarize data when they can’t act directly.</li></ul><h3>6. Prepare for the Unknown</h3><p>Agent technology is rapidly evolving. Standards, best practices, and tools are still being figured out.</p><p><strong><em>Tips:</em></strong></p><ul><li>Keep up with developments in AI agent protocols and design patterns.</li><li>Think modular: make it easy to add, remove, or update agent-friendly features.</li><li>Stay open to feedback — especially from developers or users testing agents on your product.</li></ul><h3>Wrapping Up</h3><p>Designing for <strong>Agent Experience (AX)</strong> is a natural extension of designing for human-centered UX. It’s not about replacing people — it’s about supporting how people are increasingly choosing to interact with technology.</p><p>As a designer, your job is to <strong>make the invisible visible</strong> — to structure, document, and guide experiences not just for human users, but for the intelligent systems acting on their behalf.</p><p>By embracing AX, you’re helping unlock new possibilities:</p><ul><li>Automating repetitive tasks</li><li>Reducing friction in workflows</li><li>Making your product more accessible and scalable</li></ul><p>Agents are not the future. <strong>They’re already here.</strong> And the best-designed products will be the ones ready for them.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9bebca5bfa04" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>