Search is the backbone of almost every digital product but traditional keyword search fails to capture meaning. In this session, we’ll explore how semantic search powered by embeddings and vector databases enables intelligent content discovery, personalized recommendations, and context-aware applications. We’ll go step by step: from generating embeddings with modern LLMs, storing them in a vector database, and combining them with large language models for retrieval-augmented responses.
The session includes a live demo comparing keyword search vs semantic search on the same dataset, showing how embeddings capture intent beyond exact words. Attendees will leave with a clear understanding of:
Core building blocks of semantic search (embeddings, vector DBs, RAG).

Practical tools to implement it on Google Cloud / open-source stacks.

Real-world use cases from enterprise knowledge discovery to smarter user experiences.