Germain Benard

Computer Science Engineer (ENSIMAG) and PhD in Applied Mathematics, I design and integrate end-to-end AI solutions: from data analysis and modeling to the API or application that makes the model usable.

Concretely, I work in Python (PyTorch, Hugging Face, OpenCV, LangChain) for machine learning, computer vision and LLMs/agents, and I integrate these models into products (FastAPI, Docker, React). I have notably published mobile applications, built inference APIs (YOLO, SAM2) and developed an LLM-driven generative text role-playing game.

I am particularly effective at rapid prototyping, proof of concept and integration: transforming an idea or a paper into something that works. My research background brings rigor in data and evaluation, and the habit of quickly mastering new subjects.

Available for remote work or in the Île-de-France region, full-time or part-time. Do not hesitate to contact me to discuss your needs.

Area of Expertise

Applied machine learning and computer vision: I design models, often from pre-trained models and APIs. I integrate LLMs and agents, and transform them into functional APIs and applications. Stack: Python, PyTorch, Hugging Face, FastAPI, Docker, React. I handle everything from data preparation to the delivered product, with particular ease on subjects with a strong scientific component.