LECture-bot: GenAI Course Assistant
Apr 2025 – Aug 2025 • DevOps Course Project at TUM
LECture-bot is an intelligent course material assistant designed to process uploaded lecture slides, textbooks, and notes into interactive Q&A interfaces and automated flashcards for university students.
Architecture and System Design
- RAG Subsystem: Architected the Retrieval-Augmented Generation pipeline using Python and LangChain to chunk, process, and embed uploaded course documents.
- Vector Storage: Integrated Weaviate as the vector database for efficient semantic search and similarity retrieval across course content.
- LLM Integration: Connected the microservice with both cloud-based LLM APIs (OpenAI, Gemini) and local LLM options for context-aware question answering and automatic flashcard generation.
- DevOps and CI/CD: Implemented Docker containers, automated tests with Pytest, and deployment workflows as part of the TUM DevOps course.
Links and Resources
- GitHub Repository: github.com/AET-DevOps25/team-LECture-bot