memoire-agent-V1

This is my final year's thesis project : building an AI Agent to write my final year's apprenticeship report, based on my work diary and my school's apprenticeship report guideline. It's the project I'm the most proud of because it's the essence of everything I studied and worked on combined : AI; agentic workflow; API; vector database and CI/CD pipeline. This V1 is packaging both backend and frontend leveraging the use of API, but I'm currently working on the v2 (accessible on my github) only focused on the agentic workflow (no frontend and API architecture) with enhanced AI capabilities, in terms of worflow and output.

memoire-agent-V2

The second version of my AI agent, focused on the agentic workflow with advanced capabilities. At the moment, I switched from Deepseek/ Gemini API to a 100% local solution to respect the privacy of the data I'll be using to feed the agent (work diary, which can contains private data).

optical-factory

My cleanest API infrastructure, respecting deliverables regarding precision and latency, built on a CI/CD pipeline. This was my final year's specification assignment. The project consisted in making virtual glasses try-on for online optician e-commerce. The solution we put in place can analyse the face's 3D shape from 2D (front camera) and suggest automatically a pair of glasses suited for the user