Course book
Data and Search for AI Apps
Four illustrated chapters, one for each video course. The videos build working projects; the chapters explain the ideas those projects depend on, with every calculation shown in full.
- Chapter 1Meaning as NumbersVectors and cosine similarity, tokens and embeddings, how a vector index works, and semantic search with ChromaDB.
- Chapter 2Combining Keyword and Vector SearchWhere each kind of search fails, how BM25 ranks keywords, and how Reciprocal Rank Fusion merges two result lists.
- Chapter 3Storing Data That Never Stops ArrivingWhy time-series data is different, and how chunks, columnar compression, continuous aggregates, gap filling and tiering keep it fast.
- Chapter 4Keeping Connected Agents SafeHow MCP is arranged, why agents can be misled by text, the lethal trifecta, common weaknesses and the defences that hold up.