About the role
A battery earns money by buying power when it is cheap and selling it when it is expensive. Our forecasts decide when that happens for storage sites in Texas, California and Australia. A one percent improvement in forecast error is worth real money to our customers, and it keeps more clean power on the grid.
What you will do
- Research, train and ship forecasting models for day-ahead and real-time power prices.
- Own evaluation: backtests, live monitoring and the write-ups we publish openly.
- Work with the trading software team to turn forecasts into dispatch decisions.
What you will bring
- Four or more years building ML systems that run in production, ideally with time series.
- Strong Python and a working knowledge of PyTorch or a similar framework.
- Interest in energy markets. We will teach you the rest.
Pay and benefits
Base salary plus equity. Remote anywhere in the US, with two team weeks a year in Austin. Medical, dental and vision cover, and a 401(k) match of 4%.