Applied AI / ML Engineer
Final-year AI & Data Science master's student at EPITA, graduating in 2027. My main focus is applied AI: fine-tuning and evaluating language models and building LLM-powered tools. I also work on quantitative finance and combinatorial optimization.
Bouygues Group, end-of-studies project (2026, ongoing)
AI applications for the AgiBot X2 Ultra humanoid robot, in a team of 5.
LISN / CNRS (Université Paris-Saclay), NLP research intern (Sep 2025 to Jan 2026)
Built evaluation infrastructure for language models on French biomedical text, with a multi-level, character-level scoring kit shipped as a Codabench competition. Fine-tuned CamemBERT and DrBERT, for the PARTAGES project (France 2030).
| Result | Event | What we built |
|---|---|---|
| 23rd / 110 finalists | Gradient Contest (Prologin × Mistral AI) | A game-playing bot in Python, written during a 36-hour national final and ranked in an automated tournament |
| 🥉 3rd place | HRflow.ai GenAI & HR Hackathon | Remi AI, an automated video-interview platform: conversational avatar, local transcription, LLM-as-a-judge scoring |
| 4th / 23, Smart City track | EuroTech × HKTE, Munich (EuroTech scholarship) | Manu, turning a hardware idea into a manufacturable design, stress-tested inside a learned world model |
| Top 3.8% (#845 / ~22k teams, #39 France) | IMC Prosperity 4 | Algorithmic trading: option pricing from scratch, volatility calibration, delta hedging under position limits, Monte Carlo |
- EPITA, master's in AI & Data Science, Paris (2022 to 2027)
- EPFL, first year of the CS bachelor, Lausanne (2021 to 2022)
- James Cook University, exchange semester, Singapore (2024)
- Lycée français de Prague, French baccalaureate (2014 to 2021)
- ML & NLP: Python, PyTorch, Transformers (Hugging Face), scikit-learn, pandas, NumPy, Weights & Biases, Hydra
- Optimization & quant: OR-Tools (CP-SAT), MILP (PuLP / CBC), GRASP, ALNS, option pricing, Monte Carlo
- Systems & tooling: C, C++, Java, Docker, GitLab CI, pytest, Git, Linux, Azure