This session dives deep into how Spring AI integrates with Oracle AI Vector Search to deliver high‑performance, low‑latency retrieval pipelines—while giving developers full control over embeddings, metadata, chunking, and query semantics. You’ll learn how to move beyond default settings and into deliberate, measurable optimization.
We’ll explore the internals of OracleVectorStore, including how it manages embeddings, supports hybrid search, and interacts with Oracle Database’s vector indexing capabilities. Through practical examples, you’ll see how to tune chunk sizes, adjust similarity metrics, leverage metadata filters, and optimize ingestion workflows to improve retrieval accuracy and performance.