With Databricks, you can build powerful AI Agents using foundation LLMs, Retrieval Augmented Generation (RAG), Vector Search, PDF extraction, and Mosaic AI Agent Evaluation. RAG allows you to enrich prompts with domain-specific knowledge, enabling smarter, more accurate answers—without the need to fine-tune your own models.
In this workshop, you’ll walk through an end-to-end AI agent solution on Databricks. You’ll start by exploring the AI development environment with real-time collaboration and AI assistance, then prepare and structure your data at scale. Using Unity Catalog functions and Vector Search, you’ll build a robust knowledge base and prepare documents for your agent.
From there, you’ll:
- Create and deploy your first agent with LangChain
- Deploy a real-time Q&A chatbot powered by RAG
- Scan and extract information from PDFs and documents using Databricks’ built-in ai_parse_document function
- Evaluate your agent with Mosaic AI Agent Evaluation and MLflow 3.0, and build an evaluation loop to improve performance across versions
- Monitor live agents and review production behavior
- Deploy a chatbot front-end using the Lakehouse Application
Agenda (PT)
- 11:00 AM: Introduction
- 11:05 AM: Databricks AI Overview
- 11:20 AM: Workshop
- 12:15 PM: Q&A
Duration: 1.5 hours