Tutorials

Four runnable programs, each in the repository under tutorials/, each printing its expected output. They are smoke-tested in CI, so if one does not run the build is red.

Work through them in order. Each introduces exactly one idea and depends only on the ones before it.

	What it introduces
101: Hello berth	Profile a workload, estimate it across the whole fleet, read the placement premium
201: Policies and SLOs	Objectives and constraints as plain Python; how an SLO changes the answer
301: Calibrate from Traces	Fit efficiency factors from measured traces, with held-out error and confidence intervals
401: Tail-Aware Sizing	Arrival rate plus a p99 target becomes a fleet size, with queueing headroom priced in

All four run on a laptop. No GPU, no account, no API key: the core is standard-library Python, and 301 uses the mock harness so the calibration path runs with zero accelerators.

pip install berth-placement
git clone https://github.com/reckonresearch/berth && cd berth
python tutorials/101_hello_berth.py

When you are ready to point the estimator at your own hardware rather than the simulator, go to Verify and Contribute.