The End of Search Engines? RAG Explained (Full 8-Hour Course)
Isaac Flath
Builds a search engine step by step in Python, starting with keyword search and ending with retrieval-augmented generation. The first half covers text processing, inverted indexes, TF-IDF and BM25, then embeddings, chunking and hybrid search. The second half brings in an LLM to rewrite queries, re-rank results and write answers with citations, with chapters on evaluating retrieval and on multimodal search.
What you'll learn
- Keyword search with an inverted index, TF-IDF and BM25
- Semantic search with embeddings and chunking
- Combining keyword and semantic results with reciprocal rank fusion
- Query rewriting and re-ranking with an LLM
- Measuring retrieval with precision, recall and golden datasets
- Generating answers with citations
