Burr Framework

It is a framework designed to help developers build applications that make decisions, such as chatbots, agents, and simulations, using simple Python building blocks. Burr enables users to monitor, trace, persist, and execute these applications on their own infrastructure. It is particularly well-suited for applications that utilize large language models (LLMs) and can integrate with various frameworks. Burr includes a real-time UI for tracking and monitoring system behavior, along with pluggable persisters for saving and loading application states.

Burr allows developers to express their applications as state machines, making it ideal for managing state, tracking complex decisions, incorporating human feedback, or creating idempotent, self-persisting workflows. The framework provides a simple core API and supports a wide range of applications, including time-series forecasting simulations and hyperparameter tuning. It also offers hooks and integrations for connecting with external vendors and libraries, such as Hamilton, without dictating how models are built or APIs are queried.

The framework comes with a telemetry UI that includes demo applications, such as a chatbot, to help users understand its capabilities. Burr is open-source and can be installed via PyPI, with documentation and quick-start guides available for new users. It is named after Aaron Burr, the third Vice President of the United States, and is the second open-source library released by DAGWorks, following the Hamilton library. Burr is praised for its robust state management, ease of use, and debugging capabilities, with testimonials highlighting its advantages over other frameworks like LangChain. The roadmap includes future features, and contributions from the community are encouraged.