Many streams. One flow.
Braidpipe is an async data orchestrator that fits in a single YAML file. One playbook. One Braid. Zero infrastructure.
What it does
Define where data comes from, how it transforms, and where it goes. Run it as one isolated process. No workers. No DAG servers. No containers required.
version: "1"
name: hello_braid
source:
connector: generator
config:
count: 100
prefix: "customer"
destination:
connector: stdoutInstall
pip install git+https://github.com/braidpipe/braidpipe.git@v0.1.0PyPI package coming soon.
Why it’s different
Pure choreography.
The playbook orchestrates. Your services execute. We don’t swallow your logic.
One playbook, one process.
Each Braid is isolated. If one stalls, the rest keep running.
Stateless by default.
Checkpoints and telemetry live in local SQLite. No Redis, no server.
Memory-safe.
Collections spill to disk automatically. Your RAM stays flat.
Get started
braidpipe init hello
braidpipe run ~/braids/hello/playbook.yml