Data Scientist & AI Engineer based in Paris.
I build production-grade NLP and ML systems, including fine-tuned French-language models, and turn technical work into clear value for business stakeholders.

🔭 Open to Data Scientist, AI Engineer and Data Consultant roles.

Languages: Python, SQL, DAX

ML / NLP: scikit-learn, Hugging Face Transformers (CamemBERT), time series forecasting (ARIMA, SARIMA, LSTM)

GenAI / LLM: Retrieval-Augmented Generation (RAG), LangGraph, Anthropic Claude API, Gemini API, ChromaDB, Ollama

Data viz: Power BI (star schema modeling)

Cloud / Infra: AWS (Lambda, S3, EventBridge), Docker, FastAPI, Streamlit

MLOps: MLflow, Git, GitHub Actions (CI), pytest, pre-commit, ruff, uv

sherlock: Production-grade French NLP pipeline. Built with uv, Pydantic, Typer CLI, loguru, MLflow, CamemBERT, pytest and GitHub Actions CI.

camembert-discours-politique: CamemBERT fine-tuned for multi-class classification of French political discourse. 15,000+ annotated texts, MLflow experiment tracking, thesis graded 87/100.

techradar: Serverless AWS agent (Lambda, S3, EventBridge) that scrapes and summarizes tech news with Google Gemini and LangGraph, delivered by email via SendGrid.

askmydocs: RAG assistant for querying your own documents. Built with ChromaDB, the Gemini API and a Streamlit / Docker stack, with an evaluation harness measuring hit rate and keyword recall.

HR-Dashboard: HR analytics dashboard built in Python during my Data Analyst internship at Crédit Mutuel Alliance Fédérale (HR Digital Transformation team, 50,000+ records).

newflights: Airfare optimization app. React / TypeScript / Vite frontend, FastAPI backend.