AI Engineer
Finance runs on spreadsheets, PDFs and people. We replace that with systems. When a client mandate comes in, you build what it needs: an AI operations layer for a lender, a data pipeline behind a transaction, a reporting stack, an internal tool that removes a team's worst week of the month. Then you ship it and move to the next one.
You build agentically. Models write the first draft, run the tests and do the routine. You set the spec, review the output, own the integration and answer for what runs in production. Small team, high agency, short loops from problem to working software.
You sit next to the deal team, so you learn how capital, credit and transactions actually work while you build. Over time, what you ship for clients becomes product.
- Client builds from brief to production, end to end
- Tool and model choices for each build
- The quality bar: it works, it is documented, the client can run it without you
- Turning repeat problems into internal tooling we reuse
- You have shipped things people use. Side projects count. Years do not
- You build with agents daily and have opinions on what they get wrong
- You are full-stack enough to be dangerous: backend, data, a usable front end
- You can take a vague request from a non-technical person and turn it into a spec
- Finance experience not required. Wanting to understand how money moves is
- You want the next two years to compound
Competitive salary, real ownership of what you ship, and a seat close to the deals.
Three lines on something you built and shipped, plus GitHub or CV, to info@metcap.eu, subject "AI Engineer". No cover letter.
Apply by emailWe keep application data for 12 months and delete it on request.