Job Description:
You'll join a small, focused engineering team at a fintech operating at scale globally. Your job is to take real business problems and turn them into working AI-powered solutions that people actually use.
What we're looking for:
3-6 years of solid software development (Python is your home base)
You've shipped LLM-powered tooling to production — not a side project, something people depended on
You understand the full stack: prompt engineering, tool use, agentic workflows, RAG, knowledge graphs, evals
You can wire agents into real systems (databases, CRMs, APIs) without needing everything pre-defined
You follow the AI space because you can't help it — not because it's in your job spec
Strong fit if you:
Know when to reach for RAG, when to use structured memory, when a knowledge graph fits better — and when a well-written prompt is enough
Have worked with agent frameworks and know when they're overkill
Approach agent behaviour the same way you debug code — methodically
Not the right role if you're primarily MLOps, model training, or platform-building focused.
The setup: small team, high trust, direct access to senior leadership, something meaningful in production every 4-6 weeks. Remote-friendly.
Not a researcher. Not a PoC builder. Someone who ships.