At OpenClassrooms, I completed applied data science projects spanning data exploration, predictive modeling, deep learning, and deployment. I built and evaluated Random Forest and XGBoost models to estimate building energy consumption and examine which variables influenced model performance.
I also developed an image-classification workflow using CNNs and ResNet50 transfer learning, and deployed a machine learning model through a FastAPI service and Docker. For a training recommendation assistant in France, I built a RAG proof of concept using Mistral, LangChain, and FAISS, then evaluated it with RAGAS.
France