As language models reshape how we interact with information, a new generation of AI-powered applications—language apps—is emerging. From chatbots to document search, these apps rely on vector embeddings to turn raw content into machine-understandable form. That shift opens the door to semantic search: retrieving meaning, not just matching words.

But embeddings need more than models—they need a place to live.

This talk introduces vector support in Db2, including native vector types, similarity search, and a growing set of vector operations. We’ll explore how these features enable enterprises to build scalable, secure, and intelligent applications—right where their data already lives.