AI-Ready PostgreSQL 18

Part 4

Using PostgreSQL as an AI Platform

In the fourth part of the book, you'll learn about AI-specific requirements and how they can be addressed with PostgreSQL. We'll show you how to handle natural language as input and output, and how to combine SQL data with semantic data using vectors and embeddings. pgvector is PostgreSQL's preferred tool for representing semantic information, and we'll teach you how to use it to enable semantic search, build recommendation engines, and extend simple reasoning models into multi-modal tools that support advanced Retrieval Augmented Generation (RAG) capabilities.

Separate chapters focus on integrating with Large Language Models to generate AI embeddings, building robust pipeline architectures, and leveraging ...