Key skills
PythonAILLMRAGLangChainLangGraph
About this role
Role Overview
- Production-grade AI-Agent platform using PydanticAI, LiteLLM, and Langfuse RAG pipelines with pgvector at scale
- Multi-tenant isolation, index-rebuild strategy, and hybrid search
- Multi-tenant token-quota service plus evaluation frameworks per tenant
- MCP (Model Context Protocol) tool integration as a first-class citizen
- Explainable outputs and end-to-end auditability from prompt to output for regulated industries
- Architectural decisions designed to hold up in banking, insurance, and pharmaceutical contexts
Requirements
- 5+ years of Python production engineering
- Real production LLM experience where tokens, latency, cost, and evaluation are daily realities
- RAG engineering expertise including pgvector sizing and re-ranking — not just “I know LangChain”
- Production experience with LangGraph, PydanticAI, or a comparable agent framework
- Experience with LiteLLM, Langfuse, MCP — or willingness to ramp up on them within 2 weeks
Tech Stack
- Python
Benefits
- Equity (VSOP) available — investor-backed startup for key positions
- Fully remote possible; residence in Germany required
- Team development days approx. once per month at the Munich office
- Great team culture and a lot of autonomy in your work
- You have ownership and can make a real impact