Back to Portfolio

RENCONTRE

Project Info

  • Deliverables Fullstack
  • Project Length 1.5 Months
  • View Project [RENCONTRE]

RENCONTRE: An Open Source AI-Powered Learning Platform

RENCONTRE is a full stack, open source learning platform that transforms any topic into an interactive learning map — a knowledge graph of interconnected modules and lessons that learners can explore in any direction. An AI agent researches a subject and builds an entire course in real time, while creators can also develop courses from scratch using a versatile studio featuring a live node canvas.

The platform supports a wide range of content formats, including YouTube videos, AI generated text, PDFs, quizzes, flashcards, code, and more, all with complete source attribution. Learners provide their own AI API keys, eliminating the need for platform credits. An MCP endpoint also enables external AI agents such as Claude and ChatGPT to interact directly with learning maps. The technology stack consists of a Next.js, FastAPI, and PostgreSQL monorepo deployed through Vercel and Google Cloud Run.

The broader societal impact is at the heart of RENCONTRE. Access to structured education has traditionally depended on expensive institutions, rigid schedules, and geographic accessibility. RENCONTRE aims to eliminate these barriers. Anyone with an internet connection and an interest in a subject can generate a complete, properly attributed curriculum within seconds, explore it at their own pace, and publish it for others to expand upon. Each learning map credits its sources and contributors, creating an interconnected attribution network that values shared knowledge over opaque AI generated content. In communities where access to formal education remains limited, a free and self-hostable platform capable of producing comprehensive, interactive learning experiences — rather than simple AI generated responses — can help bridge the gap between curiosity and opportunity.


RENCONTRE — mockup preview