AI Integration
I integrate LLMs, RAG pipelines, and AI toolchains into existing enterprise products. Not demos — production features behind real authentication, evaluation harnesses, and rollback paths.
I take on a small number of concurrent engagements. Each one falls into one of three shapes — chosen by the client problem, not by what is trendy this quarter. All work is delivered solo, with senior judgement, against the same engineering bar I would apply to my own product.
I integrate LLMs, RAG pipelines, and AI toolchains into existing enterprise products. Not demos — production features behind real authentication, evaluation harnesses, and rollback paths.
I migrate legacy codebases to modern stacks without breaking production. Incremental rewrites, test coverage as a gate, frozen public APIs.
I build complete products from frontend to infrastructure. One operator covers React, Elixir, Phoenix, Postgres, Docker, and the deployment pipeline.
Starting at €1,200/day
All engagements start with a paid two-week discovery sprint. From there we move to whichever shape fits the work.
Fixed-scope two-week block. Discovery, PoC, or a single production feature. Defined deliverables, agreed acceptance criteria.
Reserved capacity per month — typically 8 to 14 days. You get predictable senior delivery; I get a stable runway.
Ongoing part-time role on your engineering team, two to three days per week, three months minimum.
30-minute intro call, then a short technical exchange. We size the problem and agree what success looks like.
Written scope, deliverables, timeline, and price. One page. No proposal templates, no padding.
Two-week working block. Daily progress notes, a Friday demo, code shipped to a real branch in your repo.
Documentation, runbooks, and a recorded walkthrough. Your team owns the code on day one.
30-minute intro call. We size the problem, agree what success looks like, and decide whether to keep going.