Shubham Kumar
Product Designer · Indore, India
Systems. Interfaces.
Decisions.
Designing for decision clarity in complex and intelligent products.
- 01
AI Healthcare SaaS Platform
Arogo AI·2025·Healthcare·B2B SaaS·AIA 0-to-1 clinical product built on one hard problem: getting a time-starved doctor to treat an AI as a real clinical assistant. Sole owner across eight surfaces, from problem framing to a validated, handed-off design system.
- 02
Structural Analysis Platform
Nireekshan·2024·Civil Engineering·Structural Health Monitoring·Data VizTurning large volumes of bridge-inspection data into a decision a liable stakeholder can make in one look and defend. Three decision surfaces for a mixed-expertise, life-safety audience.
Why Being 'Right to Market' Beats Being 'First to Market' and What UX Designers Must Know
01Speed wins headlines, but getting the product right wins users. Why 'right to market' beats 'first to market', and the role UX designers play in it.
Designing Beyond Screens
02From touchscreens to voice, gesture, and ambient computing, how designers craft seamless experiences when the best interface is no interface at all.
Inclusive Design vs. Universal Design
03Two philosophies, one goal: designing for everyone. A deep look at how inclusive and universal design differ, with a Microsoft Fluent + WCAG case study.
Clinical triage dashboard
Healthcare · 2025
Underwriting decision flow
Fintech · 2025
Care-team mobile companion
Healthcare · 2024
Risk signals, visualized
Data · 2024
Design tokens & system sheet
Systems · 2024
AI compressed the time between idea and artifact. It didn’t compress the thinking. The work is still about reading people clearly, picking the right problem, and knowing when a thing is done, none of which a model has opinions about. What follows is how I move through it.
- 01
Understand
Get close enough to see what people actually do, not what they say they do, the words they reach for, the workarounds they've built, the things they've stopped complaining about because they gave up. AI chews through transcripts, finds clusters, and drafts the obvious patterns in minutes. The read on what actually matters is still mine to make.
- 02
Frame
A sharp problem statement does half the design before anyone opens Figma. I get there partly by instinct, pattern from years of watching things break the same ways, and partly by arguing with myself until the framing holds. AI is useful because it argues back: poking holes, surfacing the angle I didn't want to see. The version that survives is the one worth designing for.
- 03
Design
Most designs start as a hunch, something about the flow that feels off, or a shape the product is quietly asking to take. I follow that, then beat it into something honest. AI lets me try ten directions in the time the first one used to take, push the copy until a person could've written it, ship a working prototype before the meeting. The taste call, which one actually deserves to exist, doesn't move.
- 04
Validate
Put it in front of someone who has to decide something with it, then shut up and watch. The hesitation, the misclick, the thing they re-read twice. That's the truth, and it rarely matches the survey. AI helps me chew through feedback at volume, surface contradictions, and stress-test the edge cases I'll miss in a quiet room. What to change, and what to leave alone because the friction is the point, stays a call I make.