Data Engineer (3+ yrs) building data & ML infrastructure: pipelines, MLOps, and agentic LLM workflows. Based near Lille, open to frontalier roles (permis G).
Currently building Lazarus — a PCB defect-detection station (YOLOv11 + FastAPI + LLM diagnostic) with a swappable- detector architecture and a reproducible benchmark harness.
Toolbox — Python · SQL · Kubernetes · orchestration (Kestra) · medallion pipelines · LiteLLM / Anthropic API · local LLMs (llama.cpp, Ollama) · NixOS
Pinned projects, one story — data/ML end to end + the infra to run it:
- lazarus — computer vision + MLOps: detection, benchmarking, serving, LLM diagnostic
- notion-cv-pipeline — an agentic workflow that drafts my job applications from Notion
- homelab — GitOps K3s cluster (Traefik · MetalLB · Longhorn · Cloudflared)
- nixos-config — declarative, reproducible systems as code
Off the keyboard: triathlete, enduro rider, and relentless tinkerer.
Contact: +33.7.85.43.67.14 / pierre-louis.landouzi@epita.fr / https://www.linkedin.com/in/plland/