Pierre BAONLA BASSOM

Specializing in the development of GenAI-powered products and delivering expert training on GenAI subjects. Key Projects: • Currently building a multi-agents diagnostic tool for a client in the medical field, designed to help address challenges in underserved medical areas (medical deserts). Playground : Python, GCP, Google ADK, Open Tofu, Cloud Run, Vertex AI, Gemini • Developed and delivered a comprehensive GenAI training module for AXA GO Playground : Python, FastAPI, LLM, Agentic system, RAG, LLM Evaluation

With AI at the heart of Doctolib's strategy to transform healthcare, I worked within the ML/AI Platform team. Our goal was to deploy AI solutions to production, ensuring their scalability and reliability. • Consultation Assistant: Producing a consultation summary from a conversation. Designing the target architecture in MLP environments. Developing the backend application using Python and FastAPI. Deploying a self-hosted Speech-to-T ext model (Whisper-like). Deploying a GPT-based summary generation module. • Personal Assistant: Reducing human intervention in processing support requests. Designing the target architecture in MLP environments based on RAG, then an agentic system. Developing the backend application using Python and FastAPI. Deploying a VectorDB and an indexing pipeline. Deploying a GPT-based summary generation module. • AI Deployment Tools: Providing tools to ease the production deployment of AI applications (CI/CD, production-ready application framework, infrastructure on EKS, Python framework). • Building AI Platform around EKS Technologies: Python, AWS, EKS, ArgoCD, T erraform, Azure OpenAI, LLM, Triton Inference Server

- Involve in the payment team, in charge of the maintenance of the rating and billing platform - Research, design and implementation of scalable applications for payment ecosystem