About
I help companies build AI that people can trust and that delivers measurable value. I care about what a system decides alone, what goes back to a person, and who owns the outcome when it’s wrong.
How I work
I learn by doing. I build working prototypes myself, run experiments side by side, keep what holds up, and ask people and models where I’m wrong.
What I bring
A view across product, design and operations. I started by founding a boutique e-commerce agency, where I built and launched two online stores from scratch and developed them after launch. Then I moved into UX research and design, and from there into AI product management. At an AI startup and at a high-tech software house with clients in San Francisco, I have built AI products using AI tools and led the teams behind them.
People who work on the same problems. Through Product Cafe I know many product people in Poland, and I talk with practitioners about how they really work with AI and what they measure.
Prototypes that produce evidence. I build to find out what works, record the results and publish what they show, including what they do not prove yet.
With people
I try to turn a conflict into an experiment: what would have to be true for someone to change their mind? A founder I worked with wanted to start selling first, for sound reasons: revenue pressure and fast validation. We tested a prototype with users instead, and it showed that the key feature was different from the one we had assumed. We built that feature before selling.
I bring designers and engineers in early, and I ask one question: what is the key user problem we are trying to solve? One of my sayings is: we need the cake before we can frost it. It is about priorities, and about starting from the essence.
Background
Five years in product. I have worked and experimented with AI since 2022, and I have been building AI products since early 2024. I founded a boutique e-commerce agency and worked there as a product manager, then worked in UX research and design, and then built AI products at an AI startup and a software house.
I am strongest in product judgment, evaluation and production delivery: LLM evaluation (the translation evals), eval-set design (the planner) and human-in-the-loop design (the grant verifier).
Outside work
I co-organise Product Cafe in Poznań: I find the speakers, work out the theme of each event and host it together with a panel. I also cook.
From people who worked with me
“She actively involved engineers in the product discovery process, valued technical input, and was always open to re-evaluating priorities when new information emerged.”
“We completed everything the client asked for and even went beyond their expectations. Every milestone was delivered on time. […] Her communication was always clear and we always knew what was expected of us.”
“She has a great ability to steer the team in the right direction without letting us get lost in the details, while still making sure that the work gets done to a high standard.”
These are excerpts from recommendations on my LinkedIn profile, where you can read them in full.