Databricks Vector Search - What, Why and How

Explore the world of Databricks Vector Search in this 27-minute conference talk presented by Ankit Vij and Sonali Guleria. Discover how to enhance GenAI applications by providing the right context through semantic search, ensuring fast and accurate responses. Learn about the scalable and simple design of Databricks Vector Search, which includes data ingestion, embedding generation, and serving "search." Gain insights into developing a prototypical RAG solution and overcoming common challenges in the field, such as maintaining index health, using effective chunking strategies, improving vector search recall and precision, scaling ingestion, enhancing retrieval time, and applying governance. Equip yourself with the tools to speed up deployments to production while adhering to complex governance and compliance requirements.

Databricks Vector Search: What, Why and How