Apache Burr

Apache Burr is a self-hosted Python framework for developers building AI agents, chatbots, and other applications with branching decisions. It models application logic as a state machine, so developers can see which action ran, what changed, and why execution took a particular path. Its library and monitoring UI run on your own infrastructure, and its open-source code uses the Apache 2.0 license.

Burr's main focus is control over application state. Developers define actions and transitions with ordinary Python functions rather than a separate workflow language. The core library has no dependencies, and it works with existing model clients and frameworks instead of requiring a particular LLM provider.

The UI shows execution traces and state changes as an application runs. Saved state can live on disk, in databases, or in a custom storage backend, allowing applications to resume after an interruption. Developers can also replay earlier runs, test individual actions, and check transitions when diagnosing unexpected behavior.

For more involved workflows, Burr supports parallel actions, branching, and nested applications. Execution can pause for human input or approval. These capabilities apply to retrieval-based chatbots and multi-agent systems, as well as non-LLM applications such as forecasting simulations.

Integrations include LangChain, Apache Hamilton, Streamlit, and FastAPI. Burr also works with OpenAI and Anthropic clients, which send model requests to those external services.