This course explores vector indexing and retrieval methods for building efficient search systems with embeddings. It focuses on techniques that move beyond exact vector comparison toward scalable indexing, semantic retrieval, keyword-based retrieval, and hybrid search.

Through structured lessons and practical demonstrations, you will learn how k-nearest-neighbour search supports top-k retrieval, how FAISS indexes store and search vectors, and how dense retrieval uses embeddings for semantic matching. You will also work with sparse retrieval using BM25, combine dense and sparse rankings for hybrid search, and apply metadata-aware filtering to improve retrieval relevance. The course progresses from exact search foundations to reusable retrieval pipelines, emphasizing indexing, ranking, persistence, and document mapping. Rather than treating retrieval as a single search operation, it focuses on designing complete workflows that encode data, build indexes, retrieve relevant results, and maintain connections between vectors and source documents. By the end of this course, you will be able to: - Apply k-nearest-neighbour search and top-k retrieval for vector similarity - Build and operate FAISS indexes for efficient vector search - Implement dense semantic retrieval using Sentence Transformers - Build sparse retrieval systems using BM25-based keyword ranking - Combine dense and sparse rankings to create hybrid retrieval workflows - Design reusable retrieval pipelines with filtering, persistence, and document mapping This course is ideal for AI engineers, machine learning practitioners, developers, and professionals building semantic search and retrieval systems. A foundational understanding of Python, vector embeddings, and similarity measures is recommended; prior experience with FAISS or advanced retrieval techniques is not required. Join us to learn how to design and build efficient retrieval systems that combine vector search, keyword matching, hybrid ranking, and reusable retrieval pipelines.