I ship AI
end to end.
Product engineer building production AI surfaces — developer tooling, design systems, and the API/data layer behind them — for VC-backed founders. Active hackathon mentor on the side.
See where I can help
YEARS SHIPPING REACT7+
BASED — REMOTE-FIRSTEUProduction AI Surfaces
Developer-facing AI tooling shipped for clients in the LLM infrastructure space — at production scale, not demo scale.
0→1 Product Engineering
Greenfield Next.js apps. Stack picked from scratch — Supabase, PostgreSQL, Zustand, TanStack Query. Owned end-to-end when solo, API contracts when in a team. Vercel ship.
Design Systems & Architecture
Reusable TypeScript/React component libraries documented in Storybook. Stack selection, state-management patterns, API contracts, data-layer migrations. The boring foundations multiple product surfaces sit on.
Where I Can Help
Where I Can Help
You're shipping an LLM-powered product.
Developer-facing AI tooling and product surfaces. Built for production scale — the kind that survives real users, not just the demo.
You need a 0→1 product shipped end-to-end.
Greenfield Next.js. Stack picked from your product's actual constraints — Supabase, PostgreSQL, Zustand, TanStack Query, or whatever fits. Frontend, backend, and deploy. Owned solo or alongside your team.
Your CMS is slowing marketing down.
Headless content migration to Sanity (or similar). AI-assisted authoring flow that cuts time-to-publish from days to minutes for non-technical authors. Done in-flight without a content freeze.
Your component library isn't surviving the scale.
Reusable TypeScript/React component libraries documented in Storybook. Consumed across multiple product surfaces without forking. Theming, accessibility, and a contract that holds.
You're mid-flight and need a senior pair of hands.
End-to-end feature delivery on Next.js applications. API contracts with backend engineers when in a team, full ownership when solo. REST and WebSocket integrations. Vercel ship and post-launch iteration.
The Product
Lens.
How I think about building production AI surfaces. Same lens whether the product is greenfield or already in flight.
01. Stack from constraints.
Greenfield projects deserve a stack picked from the product's actual shape — data, latency, who'll maintain it. Never a house template.
02. Own the seams.
Solo: build the API routes too. In a team: design contracts that survive both sides shipping fast. Either way, the seam is yours.
03. Beyond the demo.
Code that survives real users — not just the live demo. Reliability, performance, and the boring parts are what ship the product.
Production code is a longer game than a demo.
What ships in week one is what runs in year three. Build for the year-three version.Community
Active mentor in the European hackathon ecosystem since 2024. Co-organiser of Hackathon for Builders (Poznań) and ADPList meetups (Poznań / Warsaw). Mentoring focus: scoping MVPs for 24–72h delivery and helping teams cut scope to ship.
- Jan 2026European Critical Infrastructure HackathonMentorGdańsk, PL
- Dec 2025HackNation PL — 1500+ participantsMentorBydgoszcz, PL
- 2025Hackathon for BuildersCo-organiserPoznań, PL
- 2024ADPList MeetupsCo-organiserPoznań / Warsaw, PL
Currently
- →7th year shipping at Tonik — building production AI surfaces for VC-backed founders.
- →Stack: TypeScript, React, Next.js, Tailwind, Supabase, PostgreSQL, Vercel.
- →Active mentor in the European hackathon ecosystem — HackNation PL, EU Critical Infrastructure.
Building AI?
Hiring a senior product engineer to ship your AI product? Need a mentor for your hackathon? Both work.