Ecosystem

Welcome to the Apache Burr Ecosystem page. This page lists all available integrations with documentation and example links.

Note

This page is a work in progress. If you have an integration or example you’d like to share, please open a PR!

---

LLM & AI Frameworks

Integration	Description	Documentation
OpenAI	Use OpenAI models (and any OpenAI API-compatible server) inside Burr actions.	Example
LangChain / LCEL	Use LangChain chains and runnables as Burr actions. Includes a custom serialization plugin to persist LangChain objects in state.	Reference | Multi-agent example | Custom serde example
Haystack	Wrap Haystack Component objects as Burr Action using HaystackAction, or convert an entire Haystack pipeline to a Burr graph.	Reference | Example
Instructor	Get structured outputs from LLMs via the Instructor library inside Burr actions.	Example

---

Orchestration & Dataflows

Integration	Description	Documentation
Hamilton	Embed Hamilton dataflows as Burr actions using the Hamilton action construct and helper functions from_state, from_value, update_state, and append_state.	Reference | Example | Multi-agent example

---

Distributed Computing

Integration	Description	Documentation
Ray	Run parallel Burr sub-applications on a Ray cluster using RayExecutor.	Reference | Example

---

State Persistence

Burr provides pluggable state persisters so your application state survives restarts and scales across services. See State Persistence for the full API reference.

Integration	Description	Documentation
SQLite	Lightweight file-based persistence. Ships with Burr (no extra install). Sync and async variants available.	Reference
PostgreSQL	Production-grade relational database persistence via psycopg2 (sync) and asyncpg (async).	Reference
Redis	In-memory key-value store persistence. Sync and async variants available.	Reference
MongoDB	Document-store persistence via pymongo.	Reference

---

Vector Stores

Integration	Description	Documentation
LanceDB	Serverless, embedded vector database. Ideal for local or cloud RAG pipelines.	RAG example
Qdrant	Scalable vector similarity search engine for conversational RAG.	Conversational RAG example

---

Observability

Integration	Description	Documentation
OpenTelemetry	Export Burr traces to any OpenTelemetry-compatible backend. Burr can also capture traces emitted within an action and forward them.	Reference | Example | Blog post
Traceloop / OpenLLMetry	AI-focused OpenTelemetry vendor. Use Traceloop’s openllmetry library together with Burr’s OpenTelemetry bridge for LLM-aware tracing.	Reference | Example | Blog post

---

Data Validation & Serialization

Integration	Description	Documentation
Pydantic	Type-check Burr state fields with Pydantic models, and serialize/deserialize state using Pydantic’s serde support.	Reference | State typing docs | Serde docs
LangChain serde	Custom serialization plugin to persist LangChain objects (messages, chains, etc.) in Burr state.	Serde docs | Example
Pandas serde	Custom serialization plugin to persist Pandas DataFrame objects in Burr state.	Serde docs
Pickle serde	Fallback serialization plugin using Python’s built-in pickle for arbitrary objects.	Serde docs

---

Web & API

Integration	Description	Documentation
FastAPI	Serve Burr applications over HTTP with streaming support via FastAPI and server-sent events (SSE).	Streaming FastAPI example | Web server example
Streamlit	Debug and visualize Burr state machines interactively in Streamlit apps using utility functions such as render_state_machine and render_explorer.	Reference | Example