ghassen nouar

I'm building production AI and LLM systems across project intelligence, media integrity, biology, green chemistry, and entrepreneurship. At DatumWave, I architect FastAPI microservices for ML inference, risk detection, forecasting, and resource optimization in Cortex PM.

I build forecasting pipelines with XGBoost and Prophet, AI risk detection using spaCy NER and FAISS similarity search, and real-time EVM/SPI/CPI dashboards over WebSockets. I also use OR-Tools CP-SAT models to optimize resource allocation.

I've published open-source PyPI tools for RAG evaluation and ML supply-chain observability, including chunkbench and Janus. My projects pair retrieval evaluation, vector search, sparse autoencoders, causal patching, Rust cores, and CI-ready automation to catch failures before they reach production.

Alongside my engineering work, I teach ML/DL workshops at Google Developers Club INSAT and mentor capstone projects. I've earned 15+ hackathon podium finishes, including six first-place wins.