Apache Incubating Project

Build reliable AI agents and applications

Apache Burr (Incubating) makes it easy to develop applications that make decisions, from simple chatbots to complex multi-agent systems. Pure Python, no magic.

0GitHub Stars
0k+PyPI Downloads
0+Discord Members

Simple, powerful Python API

Build anything from chatbots to multi-agent systems with a clean, composable interface.

chatbot.py
from burr.core import ApplicationBuilder, State, action, when

@action(reads=[], writes=["prompt", "chat_history"])
def human_input(state: State, prompt: str) -> tuple[dict, State]:
    chat_item = {"content": prompt, "role": "user"}
    return {"prompt": prompt}, state.update(
        prompt=prompt
    ).append(chat_history=chat_item)

@action(reads=["chat_history"], writes=["response", "chat_history"])
def ai_response(state: State) -> tuple[dict, State]:
    client = openai.Client()
    content = client.chat.completions.create(
        model="gpt-4o",
        messages=state["chat_history"],
    ).choices[0].message.content
    chat_item = {"content": content, "role": "assistant"}
    return {"response": content}, state.update(
        response=content
    ).append(chat_history=chat_item)

app = (
    ApplicationBuilder()
    .with_actions(human_input=human_input, ai_response=ai_response)
    .with_transitions(
        ("human_input", "ai_response"),
        ("ai_response", "human_input"),
    )
    .with_state(chat_history=[])
    .with_entrypoint("human_input")
    .build()
)
Python 3.12UTF-8burr >= 0.30

Everything you need to build AI applications

Burr provides the building blocks for reliable, observable, and testable AI-powered applications.

State Management

First-class state management with typing, persistence and serialization. Your application state is always explicit and inspectable.

Observability

Built-in UI for tracking, debugging and replaying your applications. See every decision your agent makes.

Human in the Loop

Pause, inspect and resume applications with human intervention. Keep humans in control of critical decisions.

Streaming

Native streaming support for real-time LLM token output. Build responsive UIs with streaming responses.

Parallelism

Run actions in parallel with map/reduce patterns. Scale your applications without changing the core logic.

Testing

Test your applications with snapshot-based debugging and replay. Reproduce any state your application has been in.

Works with your stack

Burr integrates with the tools and frameworks you already use. No lock-in, no wrappers.

OpenAI

LLM Provider

Anthropic

LLM Provider

Google Gemini

LLM Provider

LangChain

Framework
H

Haystack

Framework

OpenAI

LLM Provider

Anthropic

LLM Provider

Google Gemini

LLM Provider

LangChain

Framework
H

Haystack

Framework

OpenAI

LLM Provider

Anthropic

LLM Provider

Google Gemini

LLM Provider

LangChain

Framework
H

Haystack

Framework

OpenAI

LLM Provider

Anthropic

LLM Provider

Google Gemini

LLM Provider

LangChain

Framework
H

Haystack

Framework

Hamilton

Framework

FastAPI

Infrastructure

Ray

Infrastructure

OpenTelemetry

Infrastructure

Hamilton

Framework

FastAPI

Infrastructure

Ray

Infrastructure

OpenTelemetry

Infrastructure

Hamilton

Framework

FastAPI

Infrastructure

Ray

Infrastructure

OpenTelemetry

Infrastructure

Hamilton

Framework

FastAPI

Infrastructure

Ray

Infrastructure

OpenTelemetry

Infrastructure

Trusted by engineers worldwide

See what developers and teams are saying about Burr.

P
Peanut Robotics
W
Watto.ai
P
Paxton AI
P
Provectus
C
CognitiveGraphs
T
TaskHuman

After evaluating several other obfuscating LLM frame-works, their elegant yet comprehensive state management solution proved to be the powerful answer to rolling out robots driven by AI decision making.

A
Ashish Ghosh
CTO, Peanut Robotics

Of course, you can use it [LangChain], but whether it's really production-ready and improves the time from 'code-to-prod' [...], we've been doing LLM apps for two years, and the answer is no [...]. Honestly, take a look at Burr. Thank me later.

R
Reddit User
LocalLlama, Subreddit

Using Burr is a no-brainer if you want to build a modular AI application. It is so easy to build with and I especially love their UI which makes debugging, a piece of cake.

I
Ishita
Founder, Watto.ai

I just came across Burr and I'm like WOW, this seems like you guys predicted this exact need when building this. No weird esoteric concepts just because it's AI.

M
Matthew Rideout
Staff Software Engineer, Paxton AI

Burr's state management part is really helpful for creating state snapshots and build debugging, replaying and even building evaluation cases around that.

R
Rinat Gareev
Senior Solutions Architect, Provectus

I have been using Burr over the past few months, and compared to many agentic LLM platforms out there (e.g. LangChain, CrewAi, AutoGen), Burr provides a more robust framework for designing complex behaviors.

H
Hadi Nayebi
Co-founder, CognitiveGraphs

Moving from LangChain to Burr was a game-changer! It took me just a few hours to get started, compared to the days I spent navigating LangChain. I pitched Burr to my teammates and we pivoted our entire codebase to it.

A
Aditya K.
DS Architect, TaskHuman

After evaluating several other obfuscating LLM frame-works, their elegant yet comprehensive state management solution proved to be the powerful answer to rolling out robots driven by AI decision making.

A
Ashish Ghosh
CTO, Peanut Robotics

Of course, you can use it [LangChain], but whether it's really production-ready and improves the time from 'code-to-prod' [...], we've been doing LLM apps for two years, and the answer is no [...]. Honestly, take a look at Burr. Thank me later.

R
Reddit User
LocalLlama, Subreddit

Using Burr is a no-brainer if you want to build a modular AI application. It is so easy to build with and I especially love their UI which makes debugging, a piece of cake.

I
Ishita
Founder, Watto.ai

I just came across Burr and I'm like WOW, this seems like you guys predicted this exact need when building this. No weird esoteric concepts just because it's AI.

M
Matthew Rideout
Staff Software Engineer, Paxton AI

Burr's state management part is really helpful for creating state snapshots and build debugging, replaying and even building evaluation cases around that.

R
Rinat Gareev
Senior Solutions Architect, Provectus

I have been using Burr over the past few months, and compared to many agentic LLM platforms out there (e.g. LangChain, CrewAi, AutoGen), Burr provides a more robust framework for designing complex behaviors.

H
Hadi Nayebi
Co-founder, CognitiveGraphs

Moving from LangChain to Burr was a game-changer! It took me just a few hours to get started, compared to the days I spent navigating LangChain. I pitched Burr to my teammates and we pivoted our entire codebase to it.

A
Aditya K.
DS Architect, TaskHuman

After evaluating several other obfuscating LLM frame-works, their elegant yet comprehensive state management solution proved to be the powerful answer to rolling out robots driven by AI decision making.

A
Ashish Ghosh
CTO, Peanut Robotics

Of course, you can use it [LangChain], but whether it's really production-ready and improves the time from 'code-to-prod' [...], we've been doing LLM apps for two years, and the answer is no [...]. Honestly, take a look at Burr. Thank me later.

R
Reddit User
LocalLlama, Subreddit

Using Burr is a no-brainer if you want to build a modular AI application. It is so easy to build with and I especially love their UI which makes debugging, a piece of cake.

I
Ishita
Founder, Watto.ai

I just came across Burr and I'm like WOW, this seems like you guys predicted this exact need when building this. No weird esoteric concepts just because it's AI.

M
Matthew Rideout
Staff Software Engineer, Paxton AI

Burr's state management part is really helpful for creating state snapshots and build debugging, replaying and even building evaluation cases around that.

R
Rinat Gareev
Senior Solutions Architect, Provectus

I have been using Burr over the past few months, and compared to many agentic LLM platforms out there (e.g. LangChain, CrewAi, AutoGen), Burr provides a more robust framework for designing complex behaviors.

H
Hadi Nayebi
Co-founder, CognitiveGraphs

Moving from LangChain to Burr was a game-changer! It took me just a few hours to get started, compared to the days I spent navigating LangChain. I pitched Burr to my teammates and we pivoted our entire codebase to it.

A
Aditya K.
DS Architect, TaskHuman

After evaluating several other obfuscating LLM frame-works, their elegant yet comprehensive state management solution proved to be the powerful answer to rolling out robots driven by AI decision making.

A
Ashish Ghosh
CTO, Peanut Robotics

Of course, you can use it [LangChain], but whether it's really production-ready and improves the time from 'code-to-prod' [...], we've been doing LLM apps for two years, and the answer is no [...]. Honestly, take a look at Burr. Thank me later.

R
Reddit User
LocalLlama, Subreddit

Using Burr is a no-brainer if you want to build a modular AI application. It is so easy to build with and I especially love their UI which makes debugging, a piece of cake.

I
Ishita
Founder, Watto.ai

I just came across Burr and I'm like WOW, this seems like you guys predicted this exact need when building this. No weird esoteric concepts just because it's AI.

M
Matthew Rideout
Staff Software Engineer, Paxton AI

Burr's state management part is really helpful for creating state snapshots and build debugging, replaying and even building evaluation cases around that.

R
Rinat Gareev
Senior Solutions Architect, Provectus

I have been using Burr over the past few months, and compared to many agentic LLM platforms out there (e.g. LangChain, CrewAi, AutoGen), Burr provides a more robust framework for designing complex behaviors.

H
Hadi Nayebi
Co-founder, CognitiveGraphs

Moving from LangChain to Burr was a game-changer! It took me just a few hours to get started, compared to the days I spent navigating LangChain. I pitched Burr to my teammates and we pivoted our entire codebase to it.

A
Aditya K.
DS Architect, TaskHuman