Support Paul Sweeney

Modern forecasting research rarely reaches the people who'd benefit from it day-to-day. If you're not a data scientist, predicting your home's energy use means weeks of glue code and a cloud bill.

I built ML Forecast Lab (https://github.com/psweens/ml-forecast-la) to close that gap. It's a free, open-source Home Assistant add-on that benchmarks a range of ML model backends on any HA sensor, picks the winner, and publishes calibrated forecasts back. No PhD, no cloud, no GPU — runs on a Pi 5.

Coffees help me keep building it, free, for all HA users.