Avatar
🚀

Organizations

README.md

About Me

Lead Developer & AI Enthusiast based in France. I craft applications across multiple ecosystems and languages, with a strong focus on AI/LLM, software architecture, and infrastructure automation.

Homelab enthusiast, I build and experiment with local AI infrastructures, and I’ve built and automated my entire system environment with NixOS. Passionate about pushing the boundaries of what’s possible with code — from backend systems to intelligent agents.

Currently exploring the intersection of AI engineering and software development. Open to ambitious opportunities.

Skills

  • Backend: PHP, Laravel, Symfony, Go, REST APIs, GraphQL
  • Infrastructure: Docker, Proxmox, Traefik, NixOS, CI/CD
  • AI/LLM: Agentic AI CLI, Claude Code, AI Agents, Prompt Engineering
  • Interests: Homelab, Self-hosting, Open Source

Topics

  • NixOS — Declarative configuration, Nix flakes, Home Manager
  • AI & LLM — AI tools, language models, autonomous agents
  • Laravel — PHP development, packages, architecture patterns
  • Golang — CLI tools, systems programming, concurrency
  • Homelab — Self-hosted services, Docker, infrastructure
  • GitHub Stars — Trending GitHub repos, discoveries

Popular posts

  1. Hunting Tokens/sec: 4 LLM Backends, 1 Hard Ceiling (Part 2/4)
    Part 2 of 4: a benchmark journal across nixpkgs llama.cpp, upstream master, and ik_llama.cpp on Qwen3.6-27B. Six hours, four backends, all converging at 66 tok/s — and the physical reason why.

    ai llm benchmark llama-cpp qwen

  2. Speculative Decoding Meets Hybrid SSM: Why It Breaks (Part 3/4)
    Part 3 of 4: a deep-dive into why speculative decoding silently breaks (or runs anti-economically) on hybrid attention+SSM architectures like Qwen3.6, Mamba-2, and RWKV — and what would need to change upstream to fix it.

    ai llm speculative-decoding qwen mamba

  3. The NixOS Setup for llama.cpp: Declarative and Reproducible (Part 4/4)
    Part 4 of 4: the actual NixOS module, llama-pull helper, claude-code-router wiring, and one-line workflow for switching models. Five Nix files for a complete, isolated, rollback-able local LLM service.

    ai llm nixos llama-cpp systemd

  4. Why I Serve Qwen3.6 Locally on My RTX 5090 (Part 1/4)
    Part 1 of 4: motivation, hardware, and stack choices for serving Qwen3.6-27B locally on a 32 GB consumer GPU with NixOS, before any benchmarks or trade-offs kick in.

    ai llm qwen rtx5090 nixos

  5. Getting Started with Nix Flakes
    A practical introduction to Nix Flakes — the modern way to manage NixOS configurations and development environments.

    nixos nix flakes declarative

  6. Building AI Agents with Claude Code
    How to leverage Claude Code to build autonomous AI agents that publish content, review code, and manage workflows.

    ai llm claude agents automation

Less
More

    Post activity