Hybrid search in MongoDB combines full-text search (BM25 via Atlas Search) with vector search (kNN via Atlas Vector Search) to deliver results that are both keyword-precise and semantically relevant. Using the sample_mflix movies dataset, the tutorial demonstrates how to extract query vectors, run semantic searches, apply hybrid scoring with IMDb ratings, and implement Reciprocal Rank Fusion (RRF) to merge text and vector search results into a single balanced ranking. All operations run natively in MongoDB through aggregation pipelines, eliminating the need for multiple databases.