apache/burr

What is it?

Key features

• 01Build applications as state machines from plain Python functions
• 02Persist and load application state through pluggable persisters
• 03Built-in UI for real-time tracing and monitoring of execution
• 04Dependency-free core library that works with any LLM or Python framework
• 05Hooks for integrating with observability and storage vendors

Use cases

• Building stateful chatbots that remember conversation history
• Developing LLM agents with branching logic and human-in-the-loop steps
• Debugging and tracing multi-step AI workflows in a visual UI
• Running simulations or hyperparameter tuning workflows as state machines

Quick start

1. 1

   Install Burr from PyPI with the start extras

   pip install "apache-burr[start]"
2. 2

   Run the Burr UI server

   burr
3. 3

   Clone the repo and run the hello-world counter example

   git clone https://github.com/apache/burr && cd burr/examples/hello-world-counter
4. 4

   Run the counter example application

   python application.py

Things to know

• !The chatbot demo needs an OPENAI_API_KEY to actually chat, though the UI works without one.
• !The README lists async event-based orchestration and hosted execution as not yet available, and Burr Cloud is a waitlist.

In collections

Topics

FAQ

What is burr?

A Python library for building decision-making applications like chatbots and agents as state machines, with a built-in UI for tracing and monitoring. Apache Burr lets you express an application as a state machine, a graph of actions connected by transitions. Each action is a plain Python function that reads from and writes to a shared state, which makes it straightforward to manage memory, track branching decisions, insert human feedback, and resume a workflow from a saved point.

Is burr free to use? What license does it have?

burr is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.

Is burr still maintained?

Actively maintained. The last commit to burr was on Oct 21, 2026.

What can I use burr for?

Building stateful chatbots that remember conversation history Developing LLM agents with branching logic and human-in-the-loop steps Debugging and tracing multi-step AI workflows in a visual UI Running simulations or hyperparameter tuning workflows as state machines

How do I install burr?

Install Burr from PyPI with the start extras: pip install "apache-burr[start]" Run the Burr UI server: burr Clone the repo and run the hello-world counter example: git clone https://github.com/apache/burr && cd burr/examples/hello-world-counter Run the counter example application: python application.py

Where can I get burr?

The source code and installation instructions are on GitHub: https://github.com/apache/burr