Blow past benchmarks using vector search and RAG with Elasticsearch
summary
Shows a production RAG pipeline built on Elasticsearch with on‑stack embedding support (PyTorch transformers or elastic learned sparse encoder deployable with one click), hybrid ranking via reciprocal rank fusion, and an interactive serverless playground that exports runnable code. Demonstrations include semantic vs keyword searches, a Stack Overflow–backed Overflow AI prototype indexing public and private repos plus GitHub/Jira/Confluence, ANN speedups (SIMD, HW graph), and 75% RAM reduction via scalar quantization — all claimed to work at enterprise scale.
AI-written summary · about
chapters (10)
- 0:00Introduction to partnership
- 1:34Business value of search
- 7:48Implementing RAG with Elastic
- 11:21Vector search foundations
- 15:09The complete search experience
- 20:23Travel agency demo
- 24:20Stack Overflow AI overview
- 29:14Scaling enterprise knowledge
- 33:23Developer tools demonstration
- 40:12Conclusion and Q&A
related talks
45mMaking semantic search & RAG real: How to build a prod-ready app
37mBuilding advanced RAGs with Elasticsearch OpenSearch
21mSmart Search with RAG: Elasticsearch Meets Language Models
39mWhat if you could build production-ready RAG entirely in Elixir?
44mVector Search Isn’t Enough
02h 47mFrom naive to advanced RAG: the complete guide
53mBuild RAG from Scratch
19mRAG in 2025: State of the Art and the Road Forward