We produce probabilistic power forecasts for variable production and consumption for 67 hours ahead. Our forecasts provide up to 20% improvement compared to traditional forecasting methods.
By modeling both forecast and price uncertainty, we optimize the risk/profit trade-off to provide decision-support for dispatching on day-ahead and intraday markets.
Using our SaaS-platform traders and asset managers get a clear view of pre- and post-dispatch profits and imbalance costs. Reporting is easily integrated with other systems through our API.
We build decision support systems that help our customers across different power industries to get the most out of their assets.
How it works
Our system integrates data from several sources and processes it into actionable insights for our customers' daily operation.
We have build an internal terrabyte-scale data hub that continuously collects weather, production/consumption and power market data.
We produce probabilistic forecasts using advanced machine learning algorithms, such as deep learning and gradient boosting.
The probabilistic forecasts and stochastic optimization are used to calculate optimal decisions given our customers risk/profit preference.
Our products are offered as simple to use SaaS products with intuitive interface and visualizations.
Simple access to our data through our API or FTP-server.
We always run parallel systems ensuring maximum reduncancy for our customers.