Sponsored Session: AI-Powered Semantic Search with Cassandra: Today and Tomorrow
summary
Shows how to build semantic search with fastText embeddings, storing 32 paragraph chunks and handling an eight‑minute model load; with Cassandra 4 + OpenSearch with KNN plugin the index holds only IDs and vectors so OpenSearch does ANN ranking and Cassandra serves metadata, avoiding full table scans; pure Cassandra 4 without OpenSearch requires full table scans and app‑side cosine ranking (memory heavy); Cassandra 5 adds a native vector data type and an approximate nearest‑neighbor index so one query returns similarity scores and full rows.
AI-written summary · about
chapters (10)
- 0:00Semantic search in Cassandra
- 1:03Vector databases and embeddings
- 2:26Semantic search challenges
- 3:50OpenSearch with Cassandra 4
- 5:59Demo: Cassandra 4 + OpenSearch
- 14:50Cassandra 5 vector data type
- 15:54Cassandra 5 for semantic search
- 17:18Demo: Cassandra 5 vector search
- 20:54Instaclustr offerings
- 23:27Q&A
related talks
55mSupercharged Search with Semantic Search and Vector Embeddings
34mSASI, Cassandra on the full text search ride!
18mKeynote: Data = AI: Leading the Future of Gen AI with Apache Cassandra
48mUnlocking Next-gen Semantic Search With OpenSearch 3.x
11mCassandra Vector Search Performance Tuning
39mFrom Search to Insight: Leveraging OpenSearch for Scalable, AI-Driven Search Experiences
45mRevolutionizing Search with Power of Neural Sparse and Dense Vectors
28mUse OpenSearch as Your Vector Search Database To Optimize GenAI Accuracy!