<?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 Danel Roldan on Medium]]></title>
        <description><![CDATA[Stories by Danel Roldan on Medium]]></description>
        <link>https://medium.com/@dan.epecho?source=rss-1e2a0d1ddf9e------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Stories by Danel Roldan on Medium</title>
            <link>https://medium.com/@dan.epecho?source=rss-1e2a0d1ddf9e------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 02:45:27 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@dan.epecho/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 Design Problem Isn’t Always a Design Problem]]></title>
            <link>https://medium.com/@dan.epecho/the-design-problem-isnt-always-a-design-problem-57ee7107a614?source=rss-1e2a0d1ddf9e------2</link>
            <guid isPermaLink="false">https://medium.com/p/57ee7107a614</guid>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[structured-thinking]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[systems-thinking]]></category>
            <category><![CDATA[design-leadership]]></category>
            <dc:creator><![CDATA[Danel Roldan]]></dc:creator>
            <pubDate>Tue, 16 Jun 2026 14:18:27 GMT</pubDate>
            <atom:updated>2026-06-16T19:34:04.051Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Part 2 of an ongoing series. Read Part 1: “The System Is the Product.”</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fO7IJbbBRdo5QXj0cZH4Cg.png" /></figure><p>You’ve been here before. A stakeholder drops a brief in front of you. The product is broken, users are churning, and something needs to change. You open Figma. You start exploring. Three weeks later, you ship something considered and well-crafted, and nothing meaningfully improves. Not because the design was wrong. Because the design wasn’t the problem.</p><p>This happens more often than anyone admits, and at every level. But the difference between a junior designer and a senior one isn’t just craft, experience, or the complexity of problems they can hold in their heads. It’s the ability to look at what’s being called a design problem and correctly identify what kind of problem it actually is, before opening any tool at all.</p><p>That triage is the job. Everything else follows from it.</p><h3>People, Process, Product</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZEUGdTjvOmuZ-GtPwJslOQ.png" /></figure><p>Every design problem, when traced far enough, belongs to one of three categories. Sometimes more than one. But always at least one, and getting that wrong is how good design work disappears without a trace.</p><p><strong>A people problem.</strong> The team can’t agree on what the product is for, who it’s for, or what success means. Leadership says one thing; the roadmap funds another. The design brief is contradictory because the organization behind it is contradictory. When you ship something into this environment, users feel the confusion, not because the interface is wrong, but because the intent behind it was never unified. No redesign fixes this. A redesign in this context just gives the conflict a new face.</p><p><strong>A process problem.</strong> The way decisions get made is broken. Three teams are shipping to the same surface with no shared governance. Components diverge because there’s no decision rule for when to create a new one versus when to extend an existing one. Inconsistency accumulates sprint by sprint, and no component library can save a product whose process keeps undermining it. The design problem here is real, but the solution isn’t a better design. It’s a better protocol, a cleaner RACI, and a design system with teeth for enforcement rather than just documentation.</p><p><strong>A product problem.</strong> The logic itself is off. The user model is wrong. The system is solving for the wrong thing, or solving the right thing in a sequence that doesn’t match how users actually think. This is where pure design work lives, but it’s also the category that gets misapplied most often, because it’s the one designers are most trained and most eager to solve.</p><p>The mistake junior designers make, understandably, because nobody teaches this explicitly, is reaching for a design solution regardless of the problem’s category. You redesign onboarding when the real issue is that salespeople are overpromising on calls. You build a new component system when the real issue is that no one has the authority to enforce it. The design is good. The problem it’s solving isn’t the one that matters.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BvRUfvxs_KVKC7Ubt5vkVg.png" /></figure><h3>The Work Is the Diagnosis</h3><p>What senior experience gives you is the ability to triage before you design. And that triage changes everything: which stakeholders need to be in the room, what the actual success metric is, whether a design deliverable is even the right output at all.</p><p>The question isn’t “what should this look like?” It’s “which of these three constraints is the actual blocker, and what does the right intervention look like at that level?”</p><p>Sometimes you’re working on all three simultaneously because people’s misalignment is causing process dysfunction, which in turn causes product incoherence. In those cases, the job is to hold all three clearly enough to sequence the work: fix the thing that’s blocking the other things first. Don’t build the end-to-end solution on top of an unresolved people problem, because the organization will corrupt the solution before it ships.</p><p>The other thing experience teaches is which problems are yours to solve and which are impediments you have to name, document, and route to someone with the authority to fix them. A senior designer who silently absorbs a people dysfunction into their design work isn’t being strategic; they’re insulating the organization from the feedback it needs. Part of the job is making the invisible constraint visible, even when that’s uncomfortable.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YRayZJjX4XsFgvzgRPTM7Q.png" /></figure><p>Two Outcomes, Not One</p><p>Once you can reliably make that diagnosis, your solutions take on two fundamentally different forms.</p><p>The first is a quality-of-life solution. The problem is bounded. The constraint is clear. The right answer is a focused, high-quality design intervention, a workflow friction point removed, a component gap closed, a navigation pattern made coherent. This is not a lesser outcome. A quality-of-life solution delivered precisely, where it’s actually needed, has more impact than a sprawling redesign that addresses the wrong problem at scale. Most of the visible, celebrated design work in the industry is quality-of-life work done at a very high level of craft.</p><p>The second is an end-to-end solution, and it can only happen when the groundwork is right. It requires that you’ve diagnosed the people alignment, addressed or surfaced the process constraints, and have a clean product logic to build on. When those conditions are met, you’re no longer designing a screen or a flow. You’re designing the organizing logic of how the product understands itself and its users. The interface is the last step, not the first. Everything before it, the decision rules, the governance model, the component architecture, the token system, that’s the design. The screen is the output.</p><p>This is what I mean when I say the system is the product. The end-to-end solution is a system solution. And it requires the diagnosis before it’s possible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UeYOoSF6uG0RNqNr_buTyw.png" /></figure><h3>AI Will Not Change These Pillars</h3><p>Let’s address the obvious question directly, because it’s the one everyone is circling.</p><p>If AI can generate interfaces, prototype interactions, write component code, and synthesize user research, does the people/process/product framework still hold? Does the diagnostic work still belong to a human designer? Or does it get absorbed into the model along with everything else?</p><p>It doesn’t. And the reason is worth understanding precisely: it’s not just reassurance; it’s a description of what AI actually is and isn’t.</p><p>AI is extraordinary at generating solutions. It is not capable of diagnosing which problem those solutions should address. That distinction, between generation and diagnosis, is the entire gap. A model given a design brief will produce something. It has no way of knowing whether that brief is reflecting a people problem, a process problem, or a product problem. It cannot tell you that onboarding is broken because sales and product have never agreed on who the customer is. It cannot be observed that the component library keeps drifting because no one has the organizational authority to enforce it. It cannot sense that the product logic is sound, but the team dynamic is producing the wrong spec.</p><p>Those things require presence. They require reading a room, tracking contradictions across meetings, and noticing what isn’t in the brief because no one could agree on it. They require the kind of pattern recognition that only comes from having been inside broken and functional organizations and from developing the instinct to tell them apart.</p><p>This is why the people pillar, especially, is AI-proof in a way the other two are not. Process can be codified, and AI can help enforce it. Product logic can be modeled, and AI can help stress-test it. But the people dimension — the misalignments, the unspoken incentives, the organizational dysfunction that shows up as a design problem- lives entirely in the human layer. You cannot prompt your way to understanding why a team can’t ship.</p><p>What AI does change is the cost of iteration once the diagnosis is made. When you know what you’re solving for, AI compresses the distance between that clarity and a working artifact. That’s genuine and significant. But the insight has to come first. The leverage is only useful if it’s pointed at the right thing. And pointing it at the right thing is still an entirely human judgment call.</p><p>People, process, product. These are not design industry constructs. They are the irreducible categories of how organizations fail and succeed. AI will get better at everything around them. It will not replace the work of understanding which one is actually broken.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4_0IH__sAyJHGZLEjdDrkA.png" /></figure><h3>Why the Career Ladder Reorganized Around Senior</h3><p>The current market makes more sense when you understand the diagnostic frame.</p><p>The junior tier is functionally gone. Companies want Day 1 output, convinced that AI closes the gap junior hires used to require time to close. The problem with that logic is that junior work isn’t just output — it’s where you develop the triage instinct. You learn to distinguish a people problem from a product problem by getting them wrong a few times in lower-stakes situations. Cutting the junior pipeline cuts that formation, and the industry will feel it in 5 to 10 years when there aren’t enough seniors to do the diagnosis.</p><p>The principal has drifted into something adjacent to strategy consulting. At many companies, it’s where your hands leave the work, in roadmap sessions, in vision documents, and in pairing with business leadership. The people and process problems become your full-time job, but without the product grounding that makes the diagnosis credible. It’s valuable work. It’s just a different job from designing, and many designers who pursue it find themselves further from what they loved than they expected.</p><p>Senior is the level where people, process, and product are still in play simultaneously, and where your hands are still on the work. That’s the sweet spot the market is concentrating around, and it explains why more than half of hiring managers say they’re increasing demand specifically for senior design hires. Close enough to the product to diagnose it. Fluent enough in systems to solve at the right altitude. Present enough in the work to close the loop.</p><p>The designers commanding the most interesting roles right now aren’t climbing toward principal. They’re widening at the senior level, deepening their diagnostic judgment while expanding into code, systems, and strategy. Horizontal scope, not vertical title.</p><h3>What to Take From This</h3><p>Before you open Figma on your next brief, pause. Write down which of the three you think you’re actually dealing with: people, process, or product, and why. Be specific. Check yourself when the work ships. You’ll be wrong often enough at first that the pattern becomes legible.</p><p>The longer you practice this, the more clearly you’ll see which interventions actually move things and which ones just give the underlying problem a new coat of paint.</p><p>The diagnosis is the design. Everything else is execution.</p><p><em>Next: Part 3: “The Designer Who Builds.” Why code is the tool that closes the gap between diagnosis and delivery, and what the senior designer’s tech stack actually looks like in 2026.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=57ee7107a614" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The System Is The Product]]></title>
            <link>https://medium.com/@dan.epecho/the-system-is-the-product-199e44e45d87?source=rss-1e2a0d1ddf9e------2</link>
            <guid isPermaLink="false">https://medium.com/p/199e44e45d87</guid>
            <dc:creator><![CDATA[Danel Roldan]]></dc:creator>
            <pubDate>Thu, 28 May 2026 16:02:07 GMT</pubDate>
            <atom:updated>2026-06-16T11:16:41.827Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Part 1 of an ongoing series.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MmMujpfjhHBjxcldwSM_bA.png" /><figcaption>Figma + Claude + Github + Vercel = Floor Stack</figcaption></figure><h3>How Four Tools Changed the Way I Think About Product Design</h3><p><em>After 18 years in product design, I stopped collecting tools and started eliminating them. Here’s what’s left and why.</em></p><p>There’s a version of this article I could have written two years ago that would have included seventeen tools, a color-coded diagram, and a Notion template you could duplicate. It would have performed well. Designers love a good stack breakdown.</p><p>But I’m not going to write that article. Because that article is part of the problem.</p><h3>We Confused Sophistication With Accumulation</h3><p>Somewhere along the way, the design industry started measuring seniority by the size of a person’s toolkit. The more tools you knew, the more valuable you appeared. Figma, FigJam, Notion, Linear, Jira, Miro, Whimsical, Zeroheight, Storybook, Framer, Webflow, Hotjar, Maze, Amplitude, Lottie, Abstract, and whatever three new tools dropped at Config this year.</p><p>We built workflows that required twelve browser tabs, four subscriptions, and a company offsite to document. We called it a process. We called it rigor.</p><p>What it actually was was noise.</p><p>I’ve watched talented designers spend more cognitive energy maintaining their tools than using them. Updating integrations. Re-learning interfaces after major redesigns. Migrating data between platforms that no longer play nicely together. Keeping up with change logs during a period when AI is rewriting what every tool does every ninety days.</p><p>That is not a workflow. That is a second job.</p><h3>AI Changed The Calculation Permanently</h3><p>Here’s what shifted everything for me: AI didn’t just make individual tools smarter. It made most of them redundant.</p><p>When Claude can write strategy documents, generate UI prototypes, produce production-ready code, and now write directly to the Figma canvas and then pull that design back into code bidirectionally through the Figma MCP Server, the justification for a dozen single-purpose tools evaporates. Not gradually. Immediately.</p><p>The tools we adopted to bridge gaps between thinking and doing, between design and code, between prototype and product, are closing. And as they close, every tool that existed solely to bridge a gap becomes a liability rather than an asset.</p><p>The question I started asking myself wasn’t “what should I add to my stack?” It was “what can I no longer justify keeping?”</p><p>That question led me somewhere I didn’t expect.</p><h3>The Floor</h3><p>After months of ruthless elimination, I landed on four tools. Not four categories. Not four platforms with seventeen sub-tools each. Four things.</p><p><strong>Figma. Claude. GitHub. Vercel.</strong></p><p>Let me explain why each one survived and why everything else didn’t.</p><h3>Figma: The Visual Surface</h3><p>Figma isn’t in this stack because it’s the most feature-rich design tool. It’s here because it’s the shared language of the product industry. Stakeholders read it. Engineers build from it. Clients approve it. That network effect is not something you can optimize away.</p><p>But what makes Figma irreplaceable in 2026 isn’t just collaboration. It’s that Figma has absorbed an entire generation of single-purpose tools into one platform. FigJam replaced your whiteboarding tool. Figma Make replaced your prototyping tool. Figma Sites replaced your no-code publishing tool. The Figma MCP Server replaced your design-to-code handoff tool.</p><p>One subscription. One login. Every visual phase of the product lifecycle, from the first sticky note in discovery to a published marketing page.</p><p>The designers who will struggle are those who treated Figma as just a UI tool and built their workflow around everything it hadn’t yet absorbed. Those tools are gone or going.</p><h3>Claude: The Cognitive Engine</h3><p>If Figma is where design becomes visible, Claude is where everything originates.</p><p>I use Claude the way I used to use a senior strategist, a UX writer, a junior developer, and a research analyst. Not because those roles don’t matter, but because the speed at which I can now think, write, prototype, and build has fundamentally changed what one person can own in a product lifecycle.</p><p>Claude Chat handles the thinking: PRDs, research synthesis, competitive analysis, design critique, and accessibility audits. Claude Design handles rapid visual exploration before anything touches the Figma canvas. Claude Code handles the build, generating components from Figma designs, refactoring existing code, writing tests, and executing multi-step engineering tasks autonomously.</p><p>And now, through the Figma MCP Server, Claude Code writes directly to the Figma canvas and reads back from it. The round-trip is complete. Design and code are no longer separate phases with a handoff problem between them. They are the same continuous act.</p><p>That is not an incremental improvement. That is a structural change to what product design means.</p><h3>GitHub: The Infrastructure</h3><p>GitHub is here because code has to live somewhere, be versioned, reviewed, and deployed. GitHub Actions automates the pipeline, handling testing, quality checks, and deployment triggers on every commit.</p><p>No AI platform has replaced git. None will. This is load-bearing infrastructure, and it belongs in the floor of every product workflow regardless of team size, methodology, or how capable your AI tools become.</p><p>If you’re a designer who hasn’t made peace with GitHub yet, this is the year to do it. Claude Code makes it approachable in a way that nothing has before. You can describe what you want to happen and let it handle the execution. But understanding what’s happening in your repository is now a design skill, not just a developer skill.</p><h3>Vercel: The Deployment Layer</h3><p>Products have to be served to real people. Vercel does this with zero overhead and one superpower that makes it genuinely design-relevant: per-branch preview URLs.</p><p>Every time Claude Code pushes a change to GitHub, Vercel generates a live URL for that branch. That means every design iteration, not just final builds, can be reviewed in a real browser on a real device before anything merges. Design QA on live builds is no longer a phase that happens after development. It’s continuous.</p><p>Deployment should never be a bottleneck in a modern product workflow. Vercel makes sure it isn’t.</p><h3>What This Means In Practice</h3><p>The complete product lifecycle, mapped to four tools.</p><p><strong>Discover</strong> in FigJam. <strong>Define</strong> with Claude. <strong>Design</strong> in Figma. <strong>Prototype</strong> with Figma, Make, and Claude Design. <strong>Built</strong> with Claude Code through the Figma MCP Server. <strong>Ship</strong> via GitHub and Vercel. <strong>Iterate</strong> by feeding what you learn back into Claude and Figma.</p><p>No migration between platforms. No data lost in translation. No tool exists only to bridge a gap between two other tools.</p><p>Two platforms own the creative work. Figma for the visual layer, Claude for the cognitive layer. Two tools own the infrastructure. GitHub for code, Vercel for deployment.</p><p>That’s it. That is the full scope of modern product design.</p><h3>The System Is The Product. The Product Is The System.</h3><p>For most of the history of digital product design, we treated the system and the product as two separate things. The product was the output. The screens, the flows, the shipped experience that users touched. The system was the scaffolding. The tools, workflows, and handoff processes you built around the product, then tore down or ignored once it shipped.</p><p>That separation made sense when tools were isolated. When design lived in one application, code lived in another, and a human being had to manually translate between them. The system was overhead. Something you managed so it didn’t slow the product down.</p><p>That is no longer the world we are working in.</p><p>When Claude Code writes to the Figma canvas and reads back from it in a single, continuous loop, there is no longer a meaningful boundary between the design and build environments. When Figma combines your whiteboarding, prototyping, handoff, and publishing tools into one platform, the system is not supporting the product. The system has become the product. Every architectural decision you make about your stack is a product decision. Every constraint you accept in your tools becomes a constraint in what you can conceive, how fast you can move, and how well your team can collaborate around the work.</p><p>This is why the floor stack is not just a workflow preference. It is a product philosophy.</p><p>A bloated stack produces bloated products. When your tools are fragmented, your thinking becomes fragmented. When your system has seventeen handoff points, your product will have seventeen places where context gets lost, decisions get re-litigated, and quality degrades in translation. The seams in your system become the seams in your product.</p><p>A unified stack produces unified products. When your design environment and your cognitive engine are in continuous conversation, when your version control and deployment layer respond automatically to every creative decision, the product you build reflects that integrity. Fewer gaps. Fewer compromises. More of what you intended actually survives contact with production.</p><p>The most important design decisions you will make in 2026 are not about color, typography, or interaction patterns. They are about the system through which you design. Because the system is the product. And if you don’t design your system with the same intentionality you bring to your interfaces, the system will design your product for you, and not in your favor.</p><h3>The Standard I’m Proposing</h3><p>I want to be clear about what I’m not saying. I’m not saying simplicity is a virtue for its own sake. I’m not saying advanced teams don’t have legitimate reasons to add PostHog, Storybook, Linear, or other tools as their product and team scale demands.</p><p>What I am saying is that this four-tool stack is the floor, the irreducible minimum that covers the complete product lifecycle without gaps, without redundancy, and without the maintenance burden of tools that AI is actively making obsolete.</p><p>If you’re adding something to this stack, you should be able to answer one question clearly: what does this cover that Figma, Claude, GitHub, and Vercel cannot?</p><p>If the answer is “nothing, I just prefer it this way,” that’s a valid reason. Personal preference and team familiarity are real. But call it what it is.</p><p>If the answer is a genuine capability gap, add it, justify it, and own it as a deliberate choice rather than an inherited habit.</p><h3>The Designers Who Will Have The Greatest Advantage</h3><p>They won’t be the ones who know the most tools.</p><p>They’ll be the ones who internalized the pattern: discover, define, design, prototype, build, ship, learn. And developed the judgment to know which tool best serves each moment, with the confidence to remove everything that doesn’t.</p><p>They’ll also be the ones who understand that their stack is not a support structure for their work. It is their work. The system they choose to design within shapes every product that comes out of it.</p><p>AI is moving fast. Tools are changing faster than anyone can keep up with at the feature level. The only durable advantage is understanding what you’re trying to accomplish deeply enough that any tool, current or not yet invented, becomes straightforward to evaluate.</p><p>Eighteen years in, that’s the clearest thing I can tell you.</p><p>The floor is four tools. The system is the product. Build both with intention.</p><p><em>What’s in your floor stack? I’d genuinely like to know what you’ve cut and what you can’t let go of. Leave a comment below.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=199e44e45d87" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Figma Agent Didn’t Come for Your Job. It Came for Your Excuse.]]></title>
            <link>https://medium.com/@dan.epecho/the-figma-agent-didnt-come-for-your-job-it-came-for-your-excuse-e9a6fc3180ba?source=rss-1e2a0d1ddf9e------2</link>
            <guid isPermaLink="false">https://medium.com/p/e9a6fc3180ba</guid>
            <category><![CDATA[figma]]></category>
            <category><![CDATA[ai-agent]]></category>
            <category><![CDATA[strategy]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[figma-agent]]></category>
            <dc:creator><![CDATA[Danel Roldan]]></dc:creator>
            <pubDate>Thu, 21 May 2026 18:15:41 GMT</pubDate>
            <atom:updated>2026-05-21T18:15:41.654Z</atom:updated>
            <content:encoded><![CDATA[<p><em>You said you’d think more strategically when you had more time. Now you’re about to have more time.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nTTkQWB35qtTbt9WBDE4nw.png" /></figure><p>Here’s the thing nobody is saying out loud about the Figma Design Agent: it isn’t a threat to designers who think. It’s a death sentence for designers who execute.</p><p>For years, the design profession quietly sorted itself into two layers. There were designers who shaped products — who sat in the room where strategy happened, who pushed back on roadmaps, who could articulate why a decision would fail before it shipped. And there were designers who made those decisions look like screens. Both layers called themselves designers. Both layers were, by most definitions, doing their jobs. But only one of them was genuinely irreplaceable.</p><p>The Figma Agent just made the second layer a lot thinner.</p><h3>The Comfortable Lie We’ve Been Telling Ourselves</h3><p>Ask any mid-level designer why they don’t do more strategic work and you’ll hear some version of the same answer: <em>I don’t have time. I’m buried in production. The backlog never empties. By the time I’ve shipped the thing, the next thing is already waiting.</em></p><p>That answer has been true. And it has also been convenient.</p><p>Because strategy is harder than production. Strategy means having opinions that might be wrong. It means sitting in rooms with PMs and engineers and executives and defending a point of view without a Figma file to hide behind. It means being accountable for outcomes, not just outputs. Production work is hard, but the feedback loop is legible. You either built the screen or you didn’t.</p><p>The Figma Agent is about to collapse that excuse. When the agent can bulk-rename your variables, propagate design system updates across a file, explore three directions before you’ve had your morning coffee, and synthesize a round of feedback comments into actionable revisions — the production backlog compresses. The time you claimed you didn’t have starts materializing.</p><p>What you do with it will define the next chapter of your career.</p><h3>What Commoditization Actually Looks Like</h3><p>It doesn’t happen with a layoff announcement. It happens in a planning meeting when someone asks who should own the onboarding redesign, and a PM says: “can’t the agent just handle a first pass?” And they’re not entirely wrong. And nobody pushes back hard enough. And slowly the designers in the room stop being asked the interesting questions.</p><p>Commoditization in design won’t look like robots taking jobs. It will look like scope quietly shrinking. Like being looped in later. Like “just execute on this” replacing “what do you think we should do.” Like the job becoming a review layer on top of AI output rather than a generative force inside the product.</p><p>That trajectory is not inevitable. But it is the default if you don’t actively resist it.</p><h3>The Skill Gap Nobody Is Training For</h3><p>Here’s what’s strange about the current moment: design education is still optimizing for the skills the agent is best at.</p><p>Interaction patterns. Component libraries. Responsive layouts. Visual hierarchy. These are taught in bootcamps, celebrated in portfolios, evaluated in take-home assignments. They are also exactly the category of work the Figma Agent will increasingly absorb. Not because they don’t require skill — they do — but because that skill is now partially legible to a machine trained on millions of design files.</p><p>The skills the agent cannot replicate are the ones that have always been undervalued in design hiring. Can you talk to users and synthesize what you heard into a problem worth solving? Can you walk into a roadmap conversation and tell a PM that the feature they want to build is answering the wrong question? Can you look at metrics that seem fine and identify the thing the metrics aren’t measuring? Can you build organizational trust for design at a level that protects it from being deprioritized the moment the company hits a rough quarter?</p><p>These are not Figma skills. They are barely even “design” skills in the way the profession has defined design. They are the skills of someone who shapes a product, not someone who renders decisions into pixels.</p><p>The gap between those two profiles is about to get consequential.</p><h3>What Leveling Up Actually Requires</h3><p>Not a new tool. Not a course. Not a portfolio rebrand.</p><p>It requires deciding that you are a product thinker who works in a visual medium — not a visual craftsperson who attends product meetings. That sounds like semantics. It isn’t. The first identity puts you at the table when the problem is still being defined. The second puts you in the queue when the solution needs to be made shippable.</p><p>In practice, it means doing things that feel slightly outside your job description and tolerating the discomfort of that. Synthesizing user research even when research isn’t your role. Writing a design brief that articulates why a problem is worth solving before showing a single frame. Saying “I think we’re solving the wrong problem” in a roadmap meeting and being prepared to back it up. Measuring the outcomes of your work, not just the quality of your delivery.</p><p>It means making your judgment legible to the people who set strategy, so that when a decision is being made, your absence from that conversation is noticed.</p><h3>The Figma Agent Is a Test</h3><p>Not a test of your technical skill. A test of whether you were hiding behind production work, or genuinely held back by it.</p><p>The designers who treat this tool as permission to go deeper — on research, on systems thinking, on organizational influence — will find themselves more valuable than they were before it existed. The designers who use it to go faster without going further will find themselves in a shrinking lane, competing with an agent that never sleeps, never needs feedback on its craft, and gets meaningfully better every six months.</p><p>Figma shipped something that removes a constraint you said was holding you back.</p><p>Now what?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e9a6fc3180ba" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Claude’s Ecosystem is the Swiss Army Knife of Digital Product Design]]></title>
            <link>https://medium.com/@dan.epecho/how-claudes-ecosystem-is-the-swiss-army-knife-of-digital-product-design-589cb9ff7413?source=rss-1e2a0d1ddf9e------2</link>
            <guid isPermaLink="false">https://medium.com/p/589cb9ff7413</guid>
            <category><![CDATA[claude-code]]></category>
            <category><![CDATA[claude-skills]]></category>
            <category><![CDATA[claude-design]]></category>
            <category><![CDATA[claude]]></category>
            <dc:creator><![CDATA[Danel Roldan]]></dc:creator>
            <pubDate>Tue, 19 May 2026 17:37:54 GMT</pubDate>
            <atom:updated>2026-05-19T17:37:54.362Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hB8aqE8xbEENpyAEeARbFw.png" /></figure><p><strong>The Need for Versatility in Design</strong></p><p>As digital product designers, we know the value of a toolset that adapts to every stage of the journey. Today, I see Anthropic’s Claude ecosystem as a modern Swiss Army Knife—each component a precise tool, folded into a powerful, scalable system.</p><p><strong>Claude: The Core Blade</strong></p><p>At the center is Claude, a large language model that, like a trusted pocket tool, is always ready for deep reasoning. But Claude doesn’t stop there.</p><p><strong>Claude Design: Crafting Creative Precision</strong></p><p>Claude Design is like a finely tuned blade — ideal for creative refinement and visual storytelling.</p><p><strong>Claude Code: The Multi-Tool for Development</strong></p><p>Claude Code, like a multi-tool, transforms ideas into functional prototypes.</p><p><strong>Claude’s Skills: Small Attachments, Big Impact</strong></p><p>Claude’s Skills are like specialized attachments — adaptable capabilities that help summarize, map, and guide us along the way. And the Markdown files act as the pocket-sized storage, keeping context, instructions, and design decisions always within reach.</p><p><strong>Scalability and Modularity: Claude vs. Other Tools</strong></p><p>What sets Claude apart, though, is its scalability and modularity. Unlike tools like Lovable, Cursor, or Base44, which, though useful, often feel like single-purpose gadgets, Claude allows designers to outfit their workflow just like they would VS Code — extending, integrating, and scaling as needed. This modularity means we aren’t locked into isolated functions; we build a flexible, evolving design powerhouse. Lovable, Cursor, and Base44 have their merits, but they don’t scale to meet the breadth and depth that Claude does.</p><p><strong>Ready for Every Design Adventure</strong></p><p>In this way, the Claude ecosystem empowers us at every step — from ideation, to design refinement, to code handoff. Just as a Swiss Army Knife is reliable, versatile, and always within reach, Claude gives us a pocket companion for every design adventure. What tool is your Swiss Army Knife today? I’d love to hear your thoughts — drop a comment!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=589cb9ff7413" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>