Exploring Retrieval-Augmented Generation (RAG) methods with practical, reproducible experiments. We focus on building lightweight APIs, model comparisons, and exploratory demos around RAG, vector databases, embeddings, and evaluation.
Minimal RAG APIs (FastAPI + FAISS + embeddings).
Comparative experiments: Markov chains vs. LLMs, chunking strategies, rerankers.
End-to-end Spaces for testing retrieval and generation workflows.
Educational, open, and reproducible — for the community.
Minimal RAG API
Chunking, reranking, and persistence strategies (ongoing)
This org is developed and maintained by:
Naga Adithya Kaushik — ML Engineer, GenAI enthusiast.
All experiments are open-source and community-friendly.
Check out the Spaces, fork projects, and share feedback!