What you'll learn

Enterprises today can prototype AI agents quickly, but deploying them reliably, securely, and at scale remains a major challenge. In this course, Mosaic AI Agent Framework - Building and Deploying AI Agents, you’ll gain the ability to build production-ready AI agents that combine tool use, retrieval-augmented generation, and enterprise governance into a single, deployable system. First, you’ll explore the Mosaic AI Agent Framework, its core components, key features, and how it solves the real-world problems that cause most enterprise AI agent projects to stall. Next, you’ll discover how to design single-step and multi-step agents with custom tools, integrate RAG pipelines using Mosaic AI Search, and connect agents to both internal enterprise data and secure external APIs—all governed through Unity Catalog. Finally, you’ll learn how to deploy agents to Databricks Model Serving as scalable REST API endpoints, monitor production behavior with MLflow 3 tracing and inference tables, and deliver a user-friendly chat interface to end users through Databricks Apps. When you’re finished with this course, you’ll have the skills and knowledge of the Databricks Mosaic AI Agent Framework needed to build, deploy, and manage enterprise-grade AI agents that are governed, observable, and ready for production.