Build gen AI apps on Databricks

This page provides an overview of tools for building, deploying, and managing generative AI (gen AI) apps on Databricks.

- Get started: no-code GenAI
- Try AI Playground for UI-based testing and prototyping.
- Get started: MLflow 3 for GenAI
- Try MLflow for GenAI tracing, evaluation, and human feedback.

Serve and query gen AI large language models (LLMs)

Serve a curated set of gen AI models from LLM providers such as OpenAI and Anthropic and make them available through secure, scalable APIs.

Feature	Description
Foundation Models	Serve gen AI models, including open source and third-party models such as Meta Llama, Anthropic Claude, OpenAI GPT, and more.

Build and deploy your own agents, including tool-calling agents, retrieval-augmented generation apps, and multi-agent systems.

Feature	Description
AI Playground (no code)	Prototype and test AI agents in a no-code environment. Quickly experiment with agent behaviors and tool integrations before generating code for deployment.
Agent Bricks	Build and optimize domain-specific AI agent systems with a simple interface. Focus on your data and metrics while Agent Bricks streamlines implementation.
Build custom agents	Author, deploy, and evaluate agents using Python. Supports agents written with any authoring library. Integrated with MLflow Tracing. Iterate quickly using Databricks Apps.
AI agent tools	Create agent tools to query structured and unstructured data, run code, or connect to external service APIs.
MCP (Model Context Protocol)	Standardize how agents connect to data and tools with a secure, consistent interface.

Track agent performance, collect feedback, and drive quality improvements with evaluation and tracing tools.

Feature	Description
MLflow Tracing	Use MLflow Tracing for end-to-end observability. Log every step your agent takes to debug, monitor, and audit agent behavior in development and production.
Agent Evaluation	Use Agent Evaluation and MLflow to measure quality, cost, and latency. Collect feedback from stakeholders and subject matter experts through built-in review apps and use LLM judges to identify and resolve quality issues.
Monitor agents	Use the same evaluation configuration (LLM judges and custom metrics) in offline evaluation and online monitoring.