01 / AI product MVP
I handle product, UX, and code myself. You get a first version you can put in front of users, not slides about what might exist someday.
CarSense: MVP in 25 hours02 / AI in products
Assistants, recommendations, personalization. I spent 1.5 years doing this for one of the largest foodtech companies in Poland, from picking the use cases to QA after launch.
Foodtech: AI in production03 / AI agents
Agents take a task, split it between them, reach for their tools, and check each other’s work. You approve what matters.
FlowTap: agents that test appsApps I have worked on
Portfolio
AI in production at a large foodtech, my own apps in the App Store, and tools for developers. For each project I write down what I decided and why.
Products and MVPs. Each one shows the decisions, the pace, and the result.
Services
I’m most useful when an idea has to become something that works, fast: an MVP, a mobile app, an AI feature, or a tool that actually saves time.
For founders and teams who want to test an idea with real users before spending the budget on a full platform.
Proof: CarSense, an MVP in 25 hours.
More: AI product MVP →React Native and Expo. AI is part of the product here, not a sticker on the landing page.
Proof: CarSense.
More: Mobile apps with AI →I put together a full first version: frontend, backend, auth, database, payments, landing page, and docs.
Proof: TableID, FlowTap.
More: Web app / SaaS →Already have a product? I’ll add a concrete AI feature without rewriting everything from scratch.
Proof: AI in production at a foodtech, CarSense.
More: AI feature for an existing product →Systems where AI runs a series of steps, analyzes data, and hands a person the result to approve.
Proof: FlowTap, Board of Advisors.
More: Agents and automations →The project does not get passed between an analyst, a designer, and three developers. Product, UX, and code live in one head, so decisions are fast and you talk to the person who is actually building it.
How we work
One person runs the whole project. At every stage you know what you get and what it costs.
01
We sit down and take the idea apart: who it is for, what problem it solves, what has to be in the first version, and what can wait.
You getFirst-version scope
02
After the workshop we agree on the v1 scope and split it into 2-week sprints. I bill hourly on a B2B invoice. The timeline depends on the size of the app, and after every sprint you see what already works.
You getv1 scope and a sprint plan
03
I design the key screens and user flows. You review and change them now, while a change costs an hour, not a week.
You getScreen designs to approve
04
I build in 2-week sprints. After each one you hold a working version, not a status report. At the end I ship it: App Store, Google Play, or production.
You getA working version after every sprint
Available / taking on new projects
Have an idea, but not sure where to start? Let’s start with a workshop.
About
I spent 1.5 years leading AI work at one of the largest foodtech companies in Poland. After hours I ship my own apps. So I know both sides: a big product in production and an empty repo that has to become something.
I work best at the start of a project, when the scope is still moving and a few hard decisions have to be made. I design the flow and write the code myself. AI speeds up my research, implementation, and testing, but it does not think for me.
My edge is not closing the most tickets. It is deciding what to build now, what to skip, and how to get to a version people can actually try.
...
GitHub contributions in the last year
25h
from an empty repo to the CarSense MVP
1.5
years of AI in production at a foodtech
GitHub
Most of my work lives in private repositories. This is the part you can see from the outside.
Notes from building
Contact
Two sentences about the idea are enough. I’ll write back with where I would start, what I would skip, and whether I’m even the right person for it.
Email me
