<?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 Vishal phalke on Medium]]></title>
        <description><![CDATA[Stories by Vishal phalke on Medium]]></description>
        <link>https://medium.com/@vishalkphalke?source=rss-4799763c1ee3------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*lvZwQn3CagmrXgxz9kNdBw.png</url>
            <title>Stories by Vishal phalke on Medium</title>
            <link>https://medium.com/@vishalkphalke?source=rss-4799763c1ee3------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 14:36:17 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@vishalkphalke/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[How I used AI in my product design process]]></title>
            <link>https://medium.com/@vishalkphalke/how-i-used-ai-in-my-product-design-process-5a7271f3e832?source=rss-4799763c1ee3------2</link>
            <guid isPermaLink="false">https://medium.com/p/5a7271f3e832</guid>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[workflow]]></category>
            <category><![CDATA[ux-design]]></category>
            <dc:creator><![CDATA[Vishal phalke]]></dc:creator>
            <pubDate>Fri, 06 Jun 2025 15:09:45 GMT</pubDate>
            <atom:updated>2025-06-06T15:26:35.415Z</atom:updated>
            <content:encoded><![CDATA[<p>With the recent AI boom, the design world has been turned upside down — and right side up. The big debate:<br><strong>Will AI replace designers or empower them?</strong></p><figure><img alt="Pietro Schirano’s tweet which says — There is legitimately no reason to use Figma anymore." src="https://cdn-images-1.medium.com/max/601/1*GrdSrSUkUOGbt_U7JcO_-w.png" /><figcaption>Pietro Schirano challenging Figma</figcaption></figure><p>Let’s be real. Everyone’s got a hot take.<br>Some folks believe AI can design your whole product in minutes. And sure, if you’re an early-stage startup looking to launch an MVP, AI might be all you need.<br>But for companies with real users, regulated industries, or high stakes (like… saving lives?), It’s not that simple.</p><p>As designers, we’re not just moving pixels. We’re translating ambiguity into clarity, chaos into confidence. And no AI can truly replace that yet.</p><h3>The Ceiling and the Floor — and Why It Matters</h3><p>Noah Levin, VP of Design at Figma, said it best. He introduced the idea of the <strong>“Ceiling and Floor”</strong> in design:</p><ul><li><strong>Floor</strong>: The minimum skill level needed to participate in design.</li><li><strong>Ceiling</strong>: The upper limit of how far you can push creative, strategic design.</li></ul><p>AI is shifting both. It’s lowering the floor — making it easier for non-designers and junior contributors to build, test, and ship ideas. And it’s raising the ceiling — giving seasoned designers the ability to think more strategically, validate faster, and push boundaries further.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f95DUliXrtDW8x0HszOY3g.png" /><figcaption>The Ceiling and Floor in product design — by Noah Levin</figcaption></figure><p>This article is a look under the hood — a deep dive into how AI can be used to design a real-world triage platform for healthcare providers. From blank canvas to high-fidelity prototype, I didn’t just use AI to <em>make</em> things — I used it to <em>think</em>, <em>validate</em>, <em>accelerate</em>, and ultimately <em>deliver</em>.</p><p>Earlier this year, I joined a critical product initiative: building a platform that centralizes real-time alerts from wearable medical devices — think Apple Watches, fall detectors, and heart rate monitors. The mission was clear:<br> <strong>Speed up triage. Reduce noise. Save lives.</strong></p><h3><strong>Wireframe Ideation with ChatGPT</strong></h3><p>As any designer knows, starting from a blank canvas is rarely inspiring, especially when you’re designing for high-stakes environments like healthcare. So instead of jumping straight into Figma, I treated ChatGPT like a design partner.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/826/1*fj8OGjg7vVVC4k_JvVR2og.png" /><figcaption>Assigning clear quidelines to ChatGPT</figcaption></figure><p>I began by feeding it foundational research artifacts:</p><ul><li>User behavior insights on alert fatigue, triage friction, and device trust</li><li>Two key user personas</li><li>Design principles like <em>“Design for Confidence, Not Noise”</em> to ground the UX direction</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/775/1*4SPlCO3VeAfMRDkvaxTKkg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/856/1*ZxC7EFSJq2VKMbR7Lq2aHg.png" /><figcaption>(Left)Sharing user research insights with ChatGPT. (Right)Part of ChatGPT’s response</figcaption></figure><p>How was the first response it gave? Useful, but still vague. I needed precision. So I pushed further, asking ChatGPT to generate detailed wireframe descriptions <em>for each insight</em>, with a rationale for every design element.</p><p><strong>Insight-to-Interface Mapping:</strong><br>Within minutes, ChatGPT transformed abstract pain points into tangible UI concepts — each tied back to specific user needs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/798/1*xE_ivbAkLdjRcia-z-F0Lw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/814/1*iGT6VXAxbWnj6_VnPvj87g.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/789/1*2amqhEocmbh4mJisdTmvQw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/794/1*npNGCIc4ZYx7R9pc-qsaug.png" /><figcaption>Insights turned into wireframe descriptions along with rationale behind each decision</figcaption></figure><p>A few guided prompts have now been synthesized. What would typically take hours of brainstorming and iteration.</p><p><strong>The Outcome?</strong><br>I walked away with a clear layout vision for the MVP — not just what to design, but why to design it that way. Every UI component now has a validated purpose.</p><p>But design vision alone isn’t enough — it has to be shared. So, before diving into pixel-perfect work in Figma, I decided to build a fast, functional prototype using AI. The goal:<br><strong>Align cross-functional teams on the product’s experience, not just its structure.</strong></p><h3>Raising the Ceiling with AI</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/362/1*in6nveU0ein7RpZFhFZrPA.png" /><figcaption>My first thought of a triage platform given to lovable.dev as a prompt</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nQ9WVtaQY7nbhfLvaCtQ5w.png" /><figcaption>Lovable.dev’s response to my prompt</figcaption></figure><p>This prototype wasn’t perfect, not the one I had envisioned</p><p>In my first prompt to Lovable.dev, I focused only on the alert card. I needed a system, not just a screen. So I refined my prompt.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/363/1*N1Ijw_5bfkE3crBIJu4wiw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/369/1*d30JV-wwUj3V1ApwWdnvlA.png" /><figcaption>Improvised prompt for lovable.dev</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vPbGp4vrv7ngCJdIZW3c_A.png" /><figcaption>Lovable.dev’s response to my prompt</figcaption></figure><p>Within hours, I had a prototyped triage dashboard in front of me. Which adhered to usability guidelines and prioritized user experience in emergencies.</p><h3>Prototyping user flows</h3><p>Once I had clarity on the product vision and wireframe structure, the next step was to dive into specific user flows. In this section, I’ll walk you through one of the most critical use cases: escalating an active alert.<br>(<em>Interested in the other flows? Feel free to reach out: </em><a href="/service/mailto:vishalkphalke@gmail.com"><em>vishalkphalke@gmail.com</em></a>)</p><h4>Escalating an active alert.</h4><p>During user research, one insight stood out: healthcare providers often experienced emotional overload when making escalation decisions. The weight of that moment — what if I dismiss the wrong alert? — introduced hesitation and cognitive friction.</p><p>As a designer, my goal was to reduce this uncertainty by surfacing only the most critical and relevant information at the right time. In high-pressure environments like ERs, every second counts — so speed, clarity, and confidence were non-negotiable design goals.</p><p>So, to start with, I gave lovable.dev the following prompt:</p><p>“After clicking on the escalate button, the user should see current vitals of the patient, their device’s confidence score, options to choose whom they want to assign the alert to, and why they are escalating the alert”.</p><p>This is what lovable.dev showed:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BQmZaeO7Azb5MR2sHFIphw.png" /><figcaption>First iteration of escalating an alert</figcaption></figure><p>The first output used a popover modal, which seemed clean at first glance, but wasn’t suitable for this context. Here’s why:</p><ul><li><strong>Interruptive by nature</strong>: Modals block the main UI and disrupt the triage flow</li><li><strong>Loss of context</strong>: Hiding surrounding alerts reduces situational awareness</li><li><strong>Not scalable</strong>: Escalation requires rich inputs, which modals can’t accommodate well</li><li><strong>Insufficient hierarchy</strong>: Critical decisions like escalation should feel prominent, not like a subtask.</li></ul><p>Instead of a modal, I switched to a right-side sliding panel — a contextual space that keeps the primary alert feed visible while enabling deep interaction. This aligned with key UX principles:</p><ul><li><strong>Context Preservation</strong> (users stay grounded in the main flow)</li><li><strong>Progressive Disclosure</strong> (only relevant details appear when needed)</li><li><strong>Minimized Cognitive Load</strong> (no extra steps or visual clutter)</li></ul><p>Inside the panel, I wanted structured inputs like escalation targets, reason for escalation, and optional context fields, which followed <strong>Jakob’s Law</strong> by matching familiar escalation form layouts, and <strong>Hick’s Law</strong> by reducing decision friction.</p><p>To accommodate the required change, I drafted the following prompt for lovable.dev:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/414/1*-JhrCp64aD6qx9o_N2EvFg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/367/1*VjWU_jEojZ4qWba0JCtv1Q.png" /><figcaption>Escalation user flow prompt for lovable.dev</figcaption></figure><p>And the results:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*k8uxBj2h-EM-62qE5txLqg.png" /><figcaption>Right side panel for escalating an alert.</figcaption></figure><p>It was still not perfect, users had to scroll to access key inputs — a red flag in time-critical workflows. So I prompted Lovable.dev again:</p><p>“Reduce cognitive load. Improve text hierarchy. No scrolling. Make all key information visible upfront for fast action.”</p><p>The results:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kp7eAWY6Xs0Qi8FneJLahQ.png" /></figure><p>The refined version surfaced the entire decision interface in a <strong>single, scannable viewport</strong>, reducing friction and helping providers act decisively — and potentially save lives.</p><h3>Final Results</h3><p>The final prototype included:</p><ul><li>A <strong>priority-based alert inbox</strong></li><li><strong>Device trust score</strong> shown as contextual confidence</li><li>A right-side <strong>escalation panel</strong> with form logic</li><li>Real-time vitals are displayed visually</li><li>WCAG-compliant UI, validated with accessibility plugins</li></ul><p>I presented this working prototype to stakeholders within 48 hours of aligning on the “Escalate” feature. While it wasn’t the final version, it painted a clear picture of how the user could interact with the system and whether the core flows enabled the right decisions. More importantly, it saved our team days of effort, which we redirected toward usability testing and meaningful iterations.</p><h3>Why AI Didn’t Replace Design — It Amplified It</h3><p>The key to this outcome wasn’t letting AI design for me.<br>It was using AI to:</p><ul><li>Eliminate the blank canvas paralysis</li><li>Accelerate low-stakes decision-making</li><li>Act as a rapid ideation partner</li><li>Let me focus on high-impact problems</li></ul><p>AI didn’t remove the need for design thinking — it supercharged it.<br>It turned weeks of upfront framing into hours of exploration.<br>It shifted my effort from layout execution to experience orchestration.</p><p>This process made one thing clear: AI in design isn’t optional anymore — it’s the difference between shipping slowly and iterating with precision.</p><h3>Final Reflection</h3><p>This project reminded me that even with AI, good design still takes judgment.</p><p>Yes, AI helped me move faster — from blank canvas to testable prototype in hours, not days. It took on the repetitive, low-risk parts of the process so I could focus on solving real user problems. But it didn’t replace the need to make thoughtful decisions, especially in a high-stakes space like healthcare.</p><p>I still had to:</p><ul><li>Interpret user research with empathy</li><li>Decide when AI-generated outputs need a human correction or course adjustment</li><li>Push for solutions that balance clinical urgency with emotional clarity</li></ul><p>What I learned: AI is not a shortcut to good design — it’s a catalyst. It speeds up execution, but it still takes a designer to ask the right questions, frame the problem, and own the experience.</p><p>If anything, this experience made me more confident in the value of design and more aware of how powerful AI can be when used with intention.</p><p><em>Thank you for taking the time to read about how I integrated AI into my design workflow to build a real-world healthcare triage platform. If you’d like to learn more about this project, see the full prototype, or just talk design and AI, feel free to reach out to me at </em><strong><em>vishalkphalke@gmail.com</em></strong><em>. You can also explore more of my work at </em><a href="http://www.vishalphalke.com/"><strong><em>www.vishalphalke.com</em></strong></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5a7271f3e832" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[4 Tenets for budding UX Designers.]]></title>
            <link>https://medium.com/@vishalkphalke/4-tenets-for-budding-ux-designers-2020d08bfb00?source=rss-4799763c1ee3------2</link>
            <guid isPermaLink="false">https://medium.com/p/2020d08bfb00</guid>
            <category><![CDATA[ux-designer]]></category>
            <category><![CDATA[ux-principles]]></category>
            <category><![CDATA[critique]]></category>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[ideation-process]]></category>
            <dc:creator><![CDATA[Vishal phalke]]></dc:creator>
            <pubDate>Tue, 31 May 2022 06:56:21 GMT</pubDate>
            <atom:updated>2022-05-31T06:56:21.481Z</atom:updated>
            <content:encoded><![CDATA[<p>Before becoming a UX/Product Designer, I was an engineer who was enthusiastic about design but was unfamiliar with the topic of UX. I’m sure there are many more designers like myself who have only recently found the fascinating realm of UX.<br>With the intention of assisting such UX designers who are new to the profession, I derived four tenets from my personal experiences that I feel will be useful to y’all in the early stages of your UX journey.</p><h3>Different perspectives help in being innovative.</h3><blockquote>“Innovate or die” is not an idle threat. It’s the harsh reality of the modern, fast-forward economy.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-4slscehK0K9Ssw9gejUDg.jpeg" /></figure><p>A UX designer’s role is to be innovative while also assisting users by addressing their difficulties. This can be accomplished by developing a new set of eyes, not in a literal way! What I mean is looking at a problem from different perspectives is a skill that a designer must develop if they wish to assist users through their design.</p><p>While working on academic projects at IUPUI, I worked with different teams and came across some existing problem spaces. In most cases, whenever the team was handed an existing problem space, their first approach was to look for existing solutions and try improving them, which may not be the right approach always.</p><p>The point is that even when we are dealing with an existing problem space, we can approach the solution with fresh ideas and perspectives, but we restrict ourselves by looking at the current ones and try updating them. You must have a distinct point of view if you aim to provide the users with a unique product that has the capabilities to revolutionize the industry.</p><h3>The solution should always be derived from the problem.</h3><blockquote>As designers, we must move beyond being silly or ironic in order to identify a problem area and then begin developing a solution around it, rather than the other way around.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BTNXzFnUC5v6sQqHenfD5g.jpeg" /></figure><p>I worked on an academic project with a team of three HCI students during my first semester at IUPUI. We were encouraged to explore new domains, so we began exploring problem spaces without regard for boundaries or limitations. After a time, one of the team members suggested that we needed an app that would allow the user to connect with several mentors. The idea seemed very intriguing at that point, so the team agreed to work on it. No initial research was conducted to identify the problem space, and the end result was not as effective as you would think given that we worked on it for 6 weeks. As you can see, we designed a solution before recognizing a problem space in this example.</p><p>Every UX designer approaches a challenge in their own unique way. However, in general, the approach includes recognizing a problem area, empathizing with the target users, and characterizing the users’ pain points. Brainstorming, ideation, sketching, and prototyping come into the picture after that.</p><p>Starting with a solution is an inappropriate method since it completely eliminates the earliest phases of the UX process, which are critical because they aid in understanding the user and, as a consequence, aid in the design of a user-friendly product. In my perspective, a product generated from a solution-driven design is not user-centered, but rather designer-centered.</p><h3>To infinity and beyond.</h3><blockquote>“Design Fiction is the deliberate use of diegetic prototypes to suspend disbelief about change.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7UjTExeLGOqZ6DIKfTMD1Q.jpeg" /></figure><p>During the ideation phase, a designer has no limits. We are not constrained by technology or societal limitations. It is critical that we do not limit our ideas when it comes to ideation. The film industry anticipated tablets prior to the release of the iPad. This means that when seeking answers, you don’t have to think rationally or consider the limitations of current technology.</p><p>To back up this claim, I conceptualized an AI therapeutic application for one of the projects. The system was designed to help people recover from Anorexia Nervosa; an eating disorder. After doing thorough user research, observations, and interviews, I determined that therapeutic treatment can be most effective to help the target users. Because I did not consider the limitations of our present technology for this project, the solution comprised of an AI capable of performing full-length therapy sessions with the user.</p><p>The solution cannot be applied right now, but as technology evolves and Artificial Intelligence advances to the point where Siri can perform far more complicated functions than simply giving directions, we will have most of the data we need to assist people diagnosed with Anorexia Nervosa.</p><h3>Critique, don’t criticize.</h3><blockquote>“Critique is a team effort, it truly becomes valuable when we come together with the intent of understanding, identifying an opportunity, exploring, and building up those we work with.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/692/1*aBxyP01ccNgCKgHDDvxrMg.png" /></figure><p>Critique is an essential component of any design process, whether you work in a diverse team or on your own. A formal critique may help you go outside of your own thinking to make better decisions, overcome challenges, and develop your skill.</p><p>However, I believe that there is a fine line between critique and criticism that everyone must understand. I’ll use a recent example to demonstrate my point: My colleagues and I were working on the high-fidelity version of an app. Everyone came up with their own design ideas, and one of the members started criticizing designs on frivolous topics rather than presenting rationale to back up his/her point of view during the review process. Every comment was accompanied by the phrase “I don’t like it.” The evaluation quickly devolved into a heated debate. After much deliberation, we all agreed on common ground; yet, the procedure was time-consuming. It affected the team dynamics, which had a negative impact on the entire project.</p><p>This experience highlighted to me the importance of design critique and why people must recognize the difference between critique and criticism. You could have more experience or expertise than your teammates, but it won’t help you if you are unappreciative of their efforts.</p><p><em>Thank you for reading this article and I hope you found the content useful for your journey. If you have any feedback or just want to chat, you can drop a comment below, shoot me an email at vishalkphalke@gmail.com, or ping me on </em><a href="https://www.linkedin.com/in/vishal-phalke/"><em>LinkedIn</em></a><em>. Also, feel free to check out my </em><a href="https://www.vishalphalke.com/"><em>portfolio</em></a><em> for innovative projects.</em></p><h3>Citations</h3><ol><li>Tischler, L. (2001) Seven secrets to good brainstorming. Fast Company Magazine, Feb. 2001.</li><li>Mark Blythe, Kristina Andersen, Rachel Clarke, and Peter Wright. 2016. Anti-Solutionist Strategies: Seriously Silly Design Fiction. CHI 2016.</li><li>Bruce Sterling, 2012. Design fictions for Slate.</li><li>Tanner Christensen, (2016). Four Things Working at Facebook Has Taught Me About Design Critique.</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2020d08bfb00" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>