Question 1

Why do you like working at startups?

Accepted Answer

I like watching something get built from scratch. Startups usually start messy, and I enjoy helping turn that chaos into structure over time. I also really like the brainstorming and design phase, where you’re figuring out what to build and how to approach it. That part always feels the most creative to me.

Question 2

What makes you happy at work?

Accepted Answer

Honestly? When something I built actually helps someone. Not in an abstract 'this might be useful' way — but when a real person tells me it saved them time or made something less frustrating. That, and the moment a tricky AI evaluation finally starts making sense.

Question 3

What's an work-related book you'd recommend?

Accepted Answer

Building Machine Learning Powered Applications - Emmanuel Ameisen, AI Engineering: Building Applications with Foundation Models - Chip Huyen, Building Applications with AI Agents - Michae Albada, The Lean Startup - Eric Ries

Question 4

How did you get into AI?

Accepted Answer

I studied Computer Science and got pulled toward AI pretty early in my degree. Something about it just clicked — the combination of math, logic, and building things that can reason felt unlike anything else.

Question 5

What does 'product AI engineer' mean to you?

Accepted Answer

It means I care about the full thing — not just whether the model is accurate, but whether the feature actually makes sense to use. I think about the UX, I question the scope, I push back when we're optimizing the wrong thing.

Question 6

You seem to switch jobs pretty often?

Accepted Answer

I’ve always been someone who needs to feel like I’m moving forward. When I start to feel like I’m stagnating — like there’s nothing new to learn or no room to grow — I find it really hard to stay just for the sake of staying. I’d rather leave while I still respect the place than slowly become someone who stopped caring. I’m still looking for the right fit, where the team, the project, and the culture all come together. Haven’t found it yet, but I know what it feels like when things are close.

Question 7

What's something non-obvious about building with LLMs?

Accepted Answer

How much of the work is prompt engineering and evaluation, not model selection. Most people assume picking the right model is the hard part. It's not. The hard part is knowing what 'good' looks like, building a way to measure it, and iterating from there. Context engineering is also hugely underrated.

Question 8

What's the hardest part of your job?

Accepted Answer

Evals. Building AI systems is the fun part. Knowing whether they're actually working well — consistently, across edge cases, in ways that matter — that's genuinely hard. It's less glamorous than building agents, but it's what separates something you can trust from something that just demos well.

Question 9

What kind of environment do you work best in?

Accepted Answer

Small team, real problem, room to own things end-to-end. I like environments where you can move fast without five approval layers. I'm also very collaborative — I like talking through ideas, getting pushback, and iterating.

Question 10

What do you do when you're not coding?

Accepted Answer

A lot of things. I do aerial dance, which is basically gymnastics on silks. I travel whenever I can. I snowboard. I free dive. I'm very active and love doing things that keep me moving.

Question 11

Why aerial dance?

Accepted Answer

I started it a few years ago and got completely hooked. There's something about the combination of strength, coordination, and the fact that you're literally hanging in the air that demands your full attention. It's the best way I've found to not think about work. Also it looks cool, which doesn't hurt.

Question 12

What’s a skill you think more AI engineers should have?

Accepted Answer

Paying attention to details and thinking about the whole system, not just the AI part. A lot of the time the hard part is how everything fits together. User empathy is also huge - actually understanding what people need, not just what they say they want. And creativity and design thinking help a lot when you’re figuring out what to build and how it should work.

Ola Pietka — AI Engineer

Experience

Skills

Projects

Education

Contact

Ola Piętka