AI Engineering
AI Engineer
Build practical AI-powered product features, LLM integrations, intelligent workflows, and scalable AI services.
Full-timeMid–SeniorRemote (EU timezones) or KosovoFuture opportunity
This is not an open vacancy. Registering interest does not start a hiring process. We keep the note and write if the role opens.
About the role
You would add AI where it earns a place in the product: a feature that calls a model, a workflow that uses one, and the service around it. The work includes the dull failure cases — a bad answer, a slow response, and what the product does when the provider is down.
What you'll do
- Build product features that call a language model, with a defined input and output
- Write prompts and tool use that hold up on real examples, not only a demo
- Keep a small set of examples so a change can be compared with the last version
- Decide with the team which behavior should stay ordinary code
- Watch cost, latency, and which data is sent to a provider
What you bring
- You have shipped a language-model feature to users, or an internal tool people actually used
- You can build the service around the model in Python or TypeScript
- You understand structured output and retrieval, and you do not treat a prompt as a spec
- You can look at a bad answer and explain what the system did
- You are careful about which customer data is allowed to leave the product
Bonus, not required
- An evaluation set or regression check for model output
- Embeddings or a vector store inside a feature that shipped
- FastAPI or Next.js as the API in front of the model
Talk to Primeline.
Tell us what you are building. We will read it and reply with a clear view of whether we should work together.