Software Engineer - Python, backend & AI
42 graduate: Master's-level qualification (EQF level 7), Expert en architecture informatique, databases & data track. I spent 16 months at Laizy, a startup building AI assistants for businesses, working on a multi-tenant SaaS platform used in production by large companies: from the Python backend to deployment and client support.
Open to a full-time position, available now, willing to relocate (France, Switzerland or abroad).
That code is private, so here is what it did:
- LLM agents: designed the tool framework used by the platform's agents, plus 7 business tools on Microsoft Graph (mail, meetings, directory, tasks), with human validation of agent actions.
- RAG: document indexing pipeline (chunking, vector indexing in Qdrant) handling 10+ file formats, with an OCR fallback for scanned PDFs.
- Integrations: incremental SharePoint sync through Microsoft Graph, resilient to API throttling and interruptions, on 13,000+ files.
- Production support: traced 2,199 LLM API timeouts at a client to a faulty DNS resolver adding 4 s to every call, brought down to 0.04 s.
I work with coding agents (Claude Code) inside a structured process: spec, investigation, plan, implementation, review, each step documented.
- Backend: Python, FastAPI, SQLAlchemy, Pydantic, REST APIs, WebSockets
- Data: PostgreSQL, Redis, Qdrant
- AI: LLMs, RAG, agents, LangChain, LangGraph, OpenAI / Azure OpenAI
- Frontend: React, TypeScript
- Infra: Docker, Nginx, Linux, Azure, Microsoft Graph
- Also: C, C++, Bash