End-to-End Semantic Search with Apache Solr 9.8 LLM Module
summary
Apache Solr 9.8 adds an LLM module that lets Solr call external embedding services (OpenAI, Cohere, Mistral, HuggingFace via LangChain4j) to encode text at index time and query time; the release introduces knn_text_to_vector query parser and a Text To Vector Update Request Processor for in-engine vectorization, stores model configs as JSON, sends text to hosted APIs (privacy and latency implications), and plans nested/multi-vector support, quantization, and local/self‑hosted models next.
AI-written summary · about
chapters (7)
- 0:00Introduction and use case
- 4:40Vector search in Solr
- 11:38The LLM module
- 21:40Indexing with vectors
- 29:31Querying with vectors
- 32:07Future work and conclusions
- 37:59Q&A session
related talks
49mLLM—Search Engine Killer?
51mQuery Understanding With Large Language Models: Techniques and...
38mNatural Language Search-Lexical and Semantic
46mLLM—Search Engine Killer?
30mBeyond Keywords: Building an AI-Powered Search That Reads Your Mind
41mFrom Natural Language to Structured Solr Queries using LLMs
39mExperiments with OpenSearch and AI
21mBuilding RAG With OpenSearch ML Plugins: From PDFs To Voice-Enabled Search