<?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 Arjun Kani on Medium]]></title>
        <description><![CDATA[Stories by Arjun Kani on Medium]]></description>
        <link>https://medium.com/@Arjun_kani?source=rss-8265d4feb9d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*CtQA0ZJaCQbJn2pB.jpeg</url>
            <title>Stories by Arjun Kani on Medium</title>
            <link>https://medium.com/@Arjun_kani?source=rss-8265d4feb9d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 09:18:34 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@Arjun_kani/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[The Work After the Work: What Product Designers Do After Launch]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@Arjun_kani/the-work-after-the-work-what-product-designers-do-after-launch-468030fb8de4?source=rss-8265d4feb9d------2"><img src="https://cdn-images-1.medium.com/max/2600/1*4eU0WkALfqXJ4Rzkh9nD3w.jpeg" width="3727"></a></p><p class="medium-feed-snippet">There&#x2019;s a moment, right after a product ships, when everything goes quiet.</p><p class="medium-feed-link"><a href="https://medium.com/@Arjun_kani/the-work-after-the-work-what-product-designers-do-after-launch-468030fb8de4?source=rss-8265d4feb9d------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@Arjun_kani/the-work-after-the-work-what-product-designers-do-after-launch-468030fb8de4?source=rss-8265d4feb9d------2</link>
            <guid isPermaLink="false">https://medium.com/p/468030fb8de4</guid>
            <category><![CDATA[iteration]]></category>
            <category><![CDATA[post-launch]]></category>
            <category><![CDATA[product-design]]></category>
            <dc:creator><![CDATA[Arjun Kani]]></dc:creator>
            <pubDate>Wed, 24 Dec 2025 09:36:48 GMT</pubDate>
            <atom:updated>2025-12-24T09:36:48.424Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[From Data to Insight: A Practical Guide for UX Designers]]></title>
            <link>https://medium.com/design-bootcamp/from-data-to-insight-a-practical-guide-for-ux-designers-935a39c74fdb?source=rss-8265d4feb9d------2</link>
            <guid isPermaLink="false">https://medium.com/p/935a39c74fdb</guid>
            <dc:creator><![CDATA[Arjun Kani]]></dc:creator>
            <pubDate>Sun, 01 Jun 2025 09:21:38 GMT</pubDate>
            <atom:updated>2025-06-01T09:21:38.866Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yxoHmJpbOAKdzrVgauz1bQ.png" /></figure><p>Data is everywhere in UX. Clicks, scrolls, heatmaps, survey responses, user flows, you’re sitting on a goldmine. But having data isn’t the same as understanding it. The real value lies in turning that data into actionable insights that improve your product and the user experience.</p><p>This guide breaks down how to actually do that, what to look for, what to avoid, and how to make sense of the noise.</p><h3>Know the Difference: Data ≠ Insight</h3><p>Data is raw. Insights are the story the data is telling you. Think of it like this:</p><ul><li><strong>Data:</strong> 45% of users dropped off at step 3 of your sign-up flow.</li><li><strong>Insight:</strong> Step 3 asks for a phone number without explaining why users don’t trust it, so they bounce.</li></ul><p>To get from data to insight, you need to interpret context, behavior, and human motivations. That’s your job as a UX designer, not just to gather metrics, but to decode them.</p><h3>Start with a Hypothesis</h3><p>Before you dive into your analytics dashboard or user testing sessions, know what you’re looking for. Data without direction is just noise.</p><p>Ask:</p><ul><li>Why are we collecting this data?</li><li>What behavior or friction point are we investigating?</li><li>What will we do with the results?</li></ul><p>A strong hypothesis keeps your analysis focused and your conclusions relevant.</p><blockquote><em>Example: “We believe users are abandoning the checkout flow because the shipping options are unclear.”</em></blockquote><p>Now you have a lens through which to view the data.</p><h3>Look Beyond the Obvious Numbers</h3><p>Pageviews and bounce rates only tell part of the story. Useful insights come from combining multiple data sources and digging deeper.</p><p>Try:</p><ul><li><strong>Behavioral analytics</strong> (Hotjar, FullStory) to see patterns of frustration (rage clicks, hesitations).</li><li><strong>Surveys</strong> and <strong>open-text feedback</strong> to capture user intent and emotion.</li><li><strong>A/B tests</strong> to validate changes with real user impact.</li><li><strong>Journey mapping</strong> to visualize friction across steps.</li></ul><p>The insight comes from triangulating across tools and methods.</p><h3>Avoid These Common Mistakes</h3><p>Turning data into insight isn’t automatic. Watch out for these pitfalls:</p><h4>1. Mistaking correlation for causation</h4><p>Just because two things happen together doesn’t mean one causes the other. For example:</p><ul><li>Users who use dark mode stay longer.</li><li>That doesn’t mean dark mode keeps users engaged; it might just be that power users prefer dark mode.</li></ul><p>Always dig deeper before drawing conclusions.</p><h4>2. Focusing on vanity metrics</h4><p>High NPS? Great. But what are users actually doing? Prioritize behavioral data over feel-good stats unless those stats tie directly to key business goals.</p><h4>3. Ignoring qualitative context</h4><p>Quant tells you <em>what’s</em> happening. Qual tells you <em>why</em>. If 70% of users abandon at a certain step, you need interviews, open feedback, or usability testing to understand why.</p><h4>4. Synthesize, Don’t Just Report</h4><p>Dumping charts into a deck isn’t insight. Translate the findings into stories that decision-makers can act on.</p><p>Use this structure:</p><ol><li><strong>What we saw</strong> (data point)</li><li><strong>What it means</strong> (interpretation)</li><li><strong>What we recommend</strong> (design or product change)</li></ol><p><strong>Insight = data + interpretation + action.</strong></p><h4>5. Insights Should Lead to Experiments</h4><p>The goal isn’t to be right, it’s to learn and improve. Treat insights as fuel for smart design experiments.</p><p>Use them to:</p><ul><li>Prioritize backlog items.</li><li>Justify UX changes with stakeholders.</li><li>Validate or challenge assumptions.</li><li>Measure the real impact of your design decisions.</li></ul><p>Don’t just document insights, build on them.</p><h3>Keep a Living Insights Library</h3><p>Create a central place where your team can track user insights over time. This avoids repeating the same research and helps new team members ramp up quickly.</p><p>Include:</p><ul><li>What was learned</li><li>How it was learned (method/tool)</li><li>When it was learned</li><li>What actions were taken</li></ul><p>A living insights repo keeps UX design grounded in real user behavior.</p><h3>Final Thought: Data Doesn’t Replace Intuition, It Sharpens It</h3><p>UX design is still about people. Data just helps you understand them better. Use it to challenge your assumptions, guide your designs, and advocate for users more effectively.</p><p>When you combine structured analysis with empathy and creativity, you don’t just ship features, you make things better.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=935a39c74fdb" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/from-data-to-insight-a-practical-guide-for-ux-designers-935a39c74fdb">From Data to Insight: A Practical Guide for UX Designers</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Designer’s Guide to LLMs: What You Need to Know and Where to Start]]></title>
            <link>https://medium.com/design-bootcamp/a-designers-guide-to-llms-what-you-need-to-know-and-where-to-start-87dbd4f62355?source=rss-8265d4feb9d------2</link>
            <guid isPermaLink="false">https://medium.com/p/87dbd4f62355</guid>
            <dc:creator><![CDATA[Arjun Kani]]></dc:creator>
            <pubDate>Thu, 22 May 2025 16:26:15 GMT</pubDate>
            <atom:updated>2025-05-22T16:26:15.292Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*J7cp-2L0XQfw8UKcc6QWVQ.jpeg" /></figure><p>Large Language Models (LLMs) like ChatGPT, Claude, and Gemini are changing how users interact with technology. As a designer, you don’t need to be a machine learning engineer to start contributing meaningfully to LLM-powered experiences. But you do need to understand the fundamentals, and more importantly, how to design <em>for</em> and <em>with</em> these systems.</p><p>This guide is for two kinds of designers:</p><ul><li>Those who aren’t working directly with LLMs yet but want to understand them.</li><li>Those ready to design for LLM-based experiences like chatbots, smart assistants, or AI features.</li></ul><p>Let’s break it down.</p><h3>What Even Is an LLM?</h3><p>At its core, a Large Language Model is a predictive text engine trained on massive amounts of data. It doesn’t “think” or “understand” like a human; it predicts the next best word based on what came before. It’s a glorified autocomplete, but a really powerful one.</p><p>Popular LLMs:</p><ul><li><strong>OpenAI’s GPT series</strong> (ChatGPT)</li><li><strong>Anthropic’s Claude</strong></li><li><strong>Google’s Gemini</strong></li><li><strong>Meta’s LLaMA</strong></li></ul><p>What matters for designers: LLMs are not deterministic. They respond based on probability and context. The same input might not always give the same output.</p><h3>Why Should Designers Care?</h3><p>LLMs are changing interfaces. They’re pushing product design away from buttons and screens, and toward dynamic, conversational experiences. They’re also unlocking new ways to ideate, prototype, and validate ideas.</p><p>You’ll likely run into LLMs in:</p><ul><li>Conversational UIs (chatbots, voice assistants)</li><li>Smart search features</li><li>AI writing/editing tools</li><li>Adaptive help or guidance systems</li></ul><p>If you’re not designing with LLMs, you’re probably using tools powered by them — whether you realize it or not.</p><h3>The Designer’s Role in LLM Products</h3><p>You don’t need to build the model, but you do need to:</p><ul><li><strong>Define the use case.</strong> What should the AI <em>do</em>, and why?</li><li><strong>Craft prompts and user flows.</strong> LLMs respond to prompts. Designers help shape those prompts, whether they’re visible to the user or not.</li><li><strong>Design fallback and failure states.</strong> LLMs make mistakes. What happens when the model is wrong or uncertain?</li><li><strong>Set expectations.</strong> Is this AI a helper, a friend, a tool, an expert? The tone, UI, and copy must match.</li><li><strong>Prototype the experience.</strong> Figma isn’t enough here. You might need to use tools like Voiceflow, GPTs, or scripting platforms.</li></ul><h3>Key Design Principles for LLM-Driven Experiences</h3><ul><li><strong>Clarity beats cleverness.</strong> Users need to know what they can say or do. Don’t rely on them to “just try it.”</li><li><strong>Control is comfort.</strong> Let users guide the interaction. Don’t make them feel trapped in AI territory.</li><li><strong>Context is everything.</strong> LLMs thrive with clear, structured input. Give users cues or scaffolding.</li><li><strong>Transparency builds trust.</strong> If it’s AI-generated, say so. Show sources, offer options.</li><li><strong>Graceful errors are crucial.</strong> LLMs will hallucinate. Plan for it.</li></ul><h3>Tools and Resources for Designers</h3><p>If you’re just getting started, here’s how to dive in:</p><h4>No-code prototyping:</h4><ul><li><strong>Voiceflow:</strong> For conversational app design.</li><li><strong>OpenAI’s ChatGPT custom GPTs: </strong>Fast way to prototype experiences.</li><li><strong>Figma plugins: </strong>Some plugins allow prompt design/testing inside Figma.</li></ul><h4>Learn by playing:</h4><ul><li><strong>ChatGPT / Claude / Gemini: </strong>Use them. Notice how different prompts affect tone, detail, and clarity.</li><li><strong>Prompt engineering guides</strong>: Plenty of free resources on GitHub and YouTube.</li></ul><h4>Collaborate with engineers:</h4><ul><li>Share prompt ideas.</li><li>Mock up input/output examples.</li><li>Help test and iterate the LLM design is trial-and-error-heavy.</li></ul><h3>The Shift Is Already Here</h3><p>Designers used to shape static flows. Now, we shape possibilities. LLMs are not just a feature; they’re a new design material.</p><p>Understanding them isn’t optional anymore. It’s a competitive edge.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=87dbd4f62355" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/a-designers-guide-to-llms-what-you-need-to-know-and-where-to-start-87dbd4f62355">A Designer’s Guide to LLMs: What You Need to Know and Where to Start</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why UX Falls Apart After Handoff And How to Fix It]]></title>
            <link>https://medium.com/design-bootcamp/why-ux-falls-apart-after-handoff-and-how-to-fix-it-9100162f2d96?source=rss-8265d4feb9d------2</link>
            <guid isPermaLink="false">https://medium.com/p/9100162f2d96</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[visual-design]]></category>
            <dc:creator><![CDATA[Arjun Kani]]></dc:creator>
            <pubDate>Sat, 29 Mar 2025 07:33:44 GMT</pubDate>
            <atom:updated>2025-03-30T12:57:29.167Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Designer and developer in discussion" src="https://cdn-images-1.medium.com/max/1024/1*_Jjrkli_qpykul-5c7xWsA.jpeg" /></figure><p><strong>The UX handoff</strong> – when designers pass their work to developers for implementation is a crucial stage in the product development cycle. However, this is also the point where many UX designs fall apart. What looked perfect in Figma often ends up misaligned, inconsistent, or even broken once coded.</p><p>Why does this happen? Because the gap between design and development is still too wide. Designers focus on aesthetics, usability, and user experience, while developers prioritize functionality and performance. Without a seamless collaboration, the final product rarely matches the original vision.</p><p>In this blog, we’ll explore the key reasons UX deteriorates after handoff and how to prevent it. We’ll discuss the importance of working closely with developers, conducting visual QA, using Figma’s Dev Mode, and maintaining a well-defined design system.</p><h3><strong>Developers Prioritize Functionality Over UX</strong></h3><p>One of the biggest challenges is that developers are primarily concerned with making a product work rather than ensuring it looks and feels exactly as designed. This isn’t because they don’t care about UX it’s because their primary responsibility is functionality.</p><p>When deadlines are tight, developers often take shortcuts. If a button is clickable, a form submits properly, and animations don’t crash the app, they consider their job done. However, without proper oversight, small inconsistencies start to creep in wrong font sizes, incorrect spacing, missing hover states, or misaligned components.</p><p><strong>How to Fix This</strong></p><p><strong>Early Collaboration:</strong> Bring developers into the design process early. Explain why certain design choices matter for usability and user experience.</p><p><strong>Provide Context:</strong> Developers aren’t mind readers. If something is critical for UX, communicate it clearly.</p><p><strong>Encourage Ownership:</strong> If developers understand how UX affects user satisfaction, they’ll be more likely to care about maintaining design integrity.</p><h3><strong>Lack of Proper Design Documentation</strong></h3><p>A poor or nonexistent design system is another reason UX falls apart. Without clear documentation, developers often make assumptions about how components should behave. They may not have access to the correct design files, or they may reference outdated versions.</p><p>If your design system is incomplete or inconsistently followed, developers will have to “fill in the gaps” on their own. This often leads to inconsistent UI, accessibility issues, and poor scalability.</p><p><strong>How to Fix This</strong></p><p><strong>Maintain a Well-Defined Design System: </strong>A robust design system ensures consistency across screens and states. It should include typography, color palettes, spacing guidelines, button states, and interactive elements.</p><p><strong>Use Component-Based Design:</strong> Encourage developers to use reusable components instead of hardcoded elements. This prevents inconsistencies.</p><p><strong>Document Design Decisions:</strong> Annotate Figma files with clear descriptions of behavior, transitions, and edge cases.</p><h3><strong>Designers and Developers Don’t Work Closely Enough</strong></h3><p>One of the biggest mistakes in product development is treating design and development as separate phases. If designers hand off files and step back, they’re essentially hoping that developers will “get it right.”</p><p>However, design interpretation can vary. If developers don’t fully understand the intent behind certain interactions or layouts, they might implement something that technically works but doesn’t feel right to users.</p><p><strong>How to Fix This</strong></p><p><strong>Work Closely with Developers:</strong> Instead of a “handoff,” think of it as a “handover.” Continue collaborating during development.</p><p><strong>Regular Check-ins:</strong> Schedule design-development sync meetings to review progress and clarify doubts.</p><p><strong>Pair Programming Sessions:</strong> Sitting with developers while they implement key components can help catch discrepancies early.</p><h3><strong>No Visual QA (Quality Assurance)</strong></h3><p>Even with the best designs and documentation, things can still go wrong during development. Responsive behavior might break, animations may feel off, or spacing might be inconsistent. The problem? Many teams don’t do a proper visual QA before shipping.</p><p>If a designer only sees the final product right before launch, it’s too late to fix major issues without delaying the release.</p><figure><img alt="Designer doing visual qa" src="https://cdn-images-1.medium.com/max/1024/1*5KYK7gbgoVlGDY7HTh-tnQ.jpeg" /></figure><p><strong>How to Fix This</strong></p><p><strong>Conduct Visual QA Regularly:</strong> Instead of waiting until the end, check in during development to ensure the UI matches the design.</p><p><strong>Use Pixel-Perfect Comparisons:</strong> Tools like Figma’s Dev Mode, Zeplin, or Storybook can help compare coded UI with design files.</p><p><strong>Create a UX Bug List:</strong> Track visual inconsistencies just like functional bugs so they get addressed before release.</p><h3><strong>Developers Ignore (or Miss) Subtle Design Details</strong></h3><p>Designers focus on the finer details of microinteractions, hover states, and smooth animations, but these often get lost in translation. Developers may overlook these nuances, either because they aren’t explicitly defined or because they don’t see them as essential.</p><p>For example:</p><ol><li>The hover effect on a button is missing.</li><li>The loading animation is different from what was designed.</li><li>The transition between screens is too abrupt.</li></ol><p>These might seem minor, but collectively, they impact the overall user experience.</p><p><strong>How to Fix This</strong></p><p><strong>Use Dev Mode in Figma:</strong> Figma’s Dev Mode allows developers to inspect designs with exact specifications, reducing misinterpretation.</p><p><strong>Define Interaction States Clearly: </strong>Ensure all states (default, hover, active, disabled) are included in your design handoff.</p><p><strong>Provide Motion Specs:</strong> If animations are key to the UX, specify easing, duration, and behavior in the design files.</p><h3><strong>Poorly Structured or Outdated Design Files</strong></h3><p>Sometimes, UX falls apart simply because designers don’t organize their Figma files well. If developers struggle to find the right assets, they might use incorrect versions or build something from scratch instead of using the intended components.</p><p><strong>How to Fix This</strong></p><p><strong>Keep Figma Files Organized:</strong> Use clear naming conventions and structured frames so developers can easily navigate the design.</p><p><strong>Archive Old Versions:</strong> Remove outdated designs to prevent confusion.</p><p><strong>Label Components Properly:</strong> Ensure UI elements are properly categorized (e.g., buttons, inputs, cards) to align with the design system.</p><h3><strong>The Myth of “Done” After Handoff</strong></h3><p>Many teams operate under the misconception that design is “done” once it’s handed off. But design isn’t just about creating wireframes and UI it’s about ensuring those designs translate into a great user experience.</p><p>If designers disengage after the handoff, they lose the opportunity to advocate for the user and catch errors before they go live.</p><p><strong>How to Fix This</strong></p><p><strong>Stay Involved Post-Handoff:</strong> Designers should review builds, provide feedback, and help troubleshoot UX issues.</p><p><strong>Test the Live Product:</strong> Use staging environments to ensure the final implementation aligns with the intended design.</p><p><strong>Encourage a UX-First Mindset:</strong> Educate the team on why UX consistency matters, not just for aesthetics but for usability and conversion rates.</p><h3><strong>Conclusion</strong></h3><p>The UX handoff is not just a file transfer; it’s an ongoing collaboration. When designers and developers work in silos, UX inevitably suffers. But by working closely with developers, conducting visual QA, utilizing Figma’s Dev Mode, and maintaining a strong design system, you can bridge the gap and ensure your UX remains intact after handoff.</p><p>Good UX doesn’t just happen in Figma; it happens in the final product. The key is making sure every detail, from typography to animations, survives the transition from design to code.</p><p>Written by <a href="http://www.linkedin.com/in/arjun-kanithottathil">Arjun Puthiya Purayil</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9100162f2d96" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/why-ux-falls-apart-after-handoff-and-how-to-fix-it-9100162f2d96">Why UX Falls Apart After Handoff And How to Fix It</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>