Vector Search Made Simple: Getting Started With OpenSearch for AI Applications
summary
OpenSearch functions as an end to end vector database that stores dense embeddings, supports exact and approximate k-NN (HNSW, IVF), and scales to trillions of vectors while offering model connectors and integration with LangChain and LlamaIndex. Embeddings example: 1536‑dim embeddings ≈6KB each and chunking a 100 MB document into ~100k segments produces roughly 100 MB → ~100k chunks → ~600 MB embeddings. Features include text chunking, quantization to cut memory, hybrid BM25+vector scoring, and MCP support in v3.0.
AI-written summary · about
chapters (14)
- 0:00Vector search with OpenSearch
- 1:04Lexical vs semantic search
- 3:35Vector embeddings meaning
- 6:26Distance metrics & nearest neighbor
- 8:12Vector database & OpenSearch
- 9:36OpenSearch for AI apps
- 11:42Semantic search workflow
- 13:50Pre-trained & custom models
- 14:55Embedding memory demands
- 15:58Quantization for efficiency
- 17:03Hybrid lexical & semantic
- 18:51Rank normalization
- 19:32OpenSearch + AI frameworks
- 20:35Summary & getting started
related talks
30mVector Search Made Simple: Getting Started with OpenSearch for AI Applications
55mVector Search Made Simple: Getting Started with OpenSearch for AI Applications
28mUse OpenSearch as Your Vector Search Database To Optimize GenAI Accuracy!
43mOpenSearch and the Path to a Trillion-scale Vector Database
48mUnlocking Next-gen Semantic Search With OpenSearch 3.x
45mRevolutionizing Search with Power of Neural Sparse and Dense Vectors
27mMachine Learning for Advanced AI Search
55mSupercharged Search with Semantic Search and Vector Embeddings