> The complete documentation index is at [llms.txt](/llms.txt).

Bauplan is a data platform built so agents and people can work on real data without breaking it, and without managing any infrastructure.

Every change happens on a branch, every run is code you can review, and nothing reaches your production tables until you say so.

Start here →

[AI Agents](/tutorial/connect-your-agent): Use AI Agents to build pipelines, explore data, and manage your data with MCP Server and Skills.

## Get Started

- [Install Bauplan](/tutorial/installation): Set up the CLI and Python SDK and connect your API key.
- [Quick start](/tutorial/quick-start): Create a data branch, run a pipeline, and merge it to main.
- [Examples](https://github.com/BauplanLabs/bauplan/tree/main/examples): End-to-end projects you can clone and run.
- [Reference](/reference/bauplan): Every CLI command and Python SDK method.

## Go Deeper

- [Git for Data](/concepts/git-for-data): Branch, commit, and roll back your data like code.
- [Core Concepts](/concepts/projects): Projects, models, pipelines, and Iceberg tables.
- [Common Workflows](/common-scenarios/): Patterns for scheduling, parameters, and conflicts.
- [Integrations](/integrations/): Use Bauplan with orchestrators, notebooks, and BI tools.

## Watch

- [Git-for-Data and AI Agents on Production Data](https://www.youtube.com/watch?v=1x8BTeJrOKU) (5:33): Import from object storage on a branch, merge to main, and roll back a bad write.
- [Getting started with the Bauplan CLI and SDK](https://www.youtube.com/watch?v=Nd5glmOz9QA) (3:45): Install, set your API key, and work with your first data branch.
- [Platform primitives: Python, Iceberg, and Git-for-Data](https://www.youtube.com/watch?v=Di2AkSmitTc) (6:25): Write a Python model, query Iceberg tables, and version a pipeline run.

More walkthroughs and webinars on the [Bauplan YouTube channel](https://www.youtube.com/@bauplan_labs).