Matéo Le Bras SanchoFull-Stack · ML/AI Developer
CV ↓

Hi, I'm Mateo,
a Full-Stack and ML/AI Developer.

Projects

nao

nao

Open-source framework to build and deploy analytics agents — natural language queries, native data visualization, and a custom context builder for any data stack. (YC X25 · 1.3k+ stars)

TypeScriptAI SDKMCP+1
Repo
Leaffliction

Leaffliction

Computer vision app detecting plant leaf diseases across 8 categories via a CNN trained from scratch.

PythonPyTorchOpenCV+1
Repo Live
Live
Learn2Slither

Learn2Slither

RL agent trained to play Snake from scratch using Q-Learning and Deep Q-Network (DQN) with PyTorch.

PythonPyTorchQ-Learning+1
Repo Live
Live
See all projects →

About

skills & tech
C/C++TypeScriptPythonReactNext.jsNode.js
education
42 Paris2024 — present

Software Engineering

Peer-to-peer learning school — no teachers, project-based curriculum covering C/C++, Unix systems, networks, algorithms, and full-stack web.

Experience

  • AI-boosted IDE specialized for data teams at nao Labs (YC X25).
  • Full MCP integration + 2 custom MCP servers (Metabase, Fivetran); 90% reduction in token consumption.
  • TypeScript multi-agent system with specialized sub-agents, dramatically speeding up task execution.
  • Contributed to nao open-source (1.3k+ GitHub stars) — Chat with your Data via Vercel AI SDK.
  • Reduced entrepreneur drop-off during 'Mon Pass Créa' creation via an AI conversational agent.
  • Built with LangGraph for orchestration; Voice-to-Text / Text-to-Voice for natural interaction.
  • Recommendation engine for support structures, populated from a database using data collected during the exchange.
  • Deployed on Kubernetes; monitoring and CI/CD via K9S and Jenkins.
  • Developed and delivered full-stack web applications with Next.js and TypeScript.
  • End-to-end production deployment and infrastructure setup.
See all experiences →