I design fintech and B2B SaaS end to end — from problem to pixel to production.
Seven-plus years shipping to real users. Currently I'm designing a 0→1 residential PropTech platform in the UAE, using AI-assisted development to compress the design-to-prototype loop into something close to the speed of thought.
Before that, Qashio — where I cut KYB from 50 minutes to 15 and lifted activation by 22%. Before that, the JazzCash redesign — a team effort at IBM/Systems Limited, on a wallet serving ~9 million monthly users at the time — taught me that scale is a design constraint, not a design metric.
I care about the operational reality of the teams that live with software. Screens are the residue. Systems, language, and honest handoff are the actual work.
- Based
- Dubai, UAE
- Availability
- Select roles · remote / hybrid
- Experience
- 7+ years
- Users shipped to
- ~9M MAU (at the time)
- Education
- BS Computer Science (Hons), Forman Christian College
Where I've shipped.
- 2026 — Present
UAE PropTech Startup
Product Designer (Contract)Currently designing a 0→1 residential PropTech platform. AI-assisted prototyping, validated with property managers before each iteration.
- 2024 — 2026
Qashio
Product DesignerRedesigned KYB and auto-onboarding (50→15 min, +22% activation). Owned invoice & receipt system (80% adoption in month one, −40% duplicates). Built scalable design system across mobile and web.
- 2023 — 2024
WalletWise
Founding Product DesignerSole designer on a multi-currency consumer expense app. IA, personas, flows, and a component library from scratch. −40% task steps vs. competitor benchmarks.
- 2021 — 2023
Dubizzle Labs
Product DesignerRedesigned the real estate whiteboard platform used by 200+ agencies and 50,000+ agents. Designed Dubizzle Jobs from scratch. Contributed to BayutPro mobile.
- 2019 — 2021
Systems Limited
User Experience DesignerOne of the designers on the IBM/Systems Limited redesign of JazzCash (~9M monthly active users at the time). Send-money flow 5→3 steps, +15% success rate. Bulk payments +20% adoption in three months. Designed CRM for MattressFirm (24,000 stores).
How I work, in full.
I design for the person who isn't me.
Different ages, languages, cultures, levels of literacy, and mental models. I get close to the people nearest the problem and design around what they actually know, notice, and expect.
On the JazzCash redesign, the people taking payments were not the people writing the requirements. Reading ability, language and comfort with a phone varied enormously, so recognition — icons, position, repetition — had to carry more of the flow than words did. I read the funnel to find where people stopped, then designed the moment just before that.
I don't start with screens.
I start with the problem, and bring product, engineering, and the people closest to it in early — so we decide what should change before deciding how it should look.
In practice that means a shared read of the evidence before any Figma file exists: what people are actually doing, where they give up, what the business needs to be true. Screens come after the decision, not instead of it.
AI needs clear boundaries.
I use AI to explore and prototype faster, but never to outsource judgment. In the product, I design when AI should ask, act, explain, or hand control back.
The interesting design work in an AI feature is rarely the output — it's the relationship. What the system is confident about, what it shows its reasoning for, and what it never does without a person saying yes.
Shipping is part of the design work.
I stay close through build, QA, launch, and what the numbers say after. If something moves the wrong way, I help fix it.
Design ships when it's in production, instrumented, and holding up under real load. I sit with engineering through the sprint, check the funnel a week later, and ship a patch when the number moves the wrong way.
A design system earns its weight on Monday morning.
Tokens and components are worth what they save the team on an ordinary working day — not what they look like in a showcase file. I ship the primitives an engineer can build against in an afternoon, and the documentation that ends the sixteenth debate about button padding.
That means naming that matches how the team already talks, states covered before they're needed, and adoption measured by what engineering actually uses rather than what exists in the library. A system nobody reaches for is a very tidy liability.