Forecasts
Short/long-term demand & PV forecasts with calendar/temperature effects and uncertainty bands.
An energy plan depends on what demand and generation will do next. Forecast demand and PV, investigate anomalies and compare cost scenarios with explainable models; send the results to the teams and systems that act on them.
Work from a forecast to the reason behind it, then test the operational or tariff decision it is intended to support.
Short/long-term demand & PV forecasts with calendar/temperature effects and uncertainty bands.
Find outliers, missing/late reads and drift. Each alert includes cause hypotheses and context.
Tariff/TOU scenarios, storage scheduling and curtailment what-ifs with expected impact.
Feature importance, SHAP-like summaries and confidence scores for every prediction.
Push forecasts & alerts via webhooks, query KPIs and export batches to BI/finance.
Ready views for communities, utilities and finance with exportable, auditable reports.
The analytics layer consumes normalized readings and contextual data. Its output is a forecast, alert or scenario for an operational decision; settlement and billing remain downstream consumers.
Combine Gateway or Platform readings with weather and calendar data. Quality checks and lineage let the team examine which inputs a prediction used.
Use AutoML and curated model templates with backtests, retraining schedules and drift alerts. Review model performance as the underlying consumption or generation pattern changes.
Pass forecasts and alerts through APIs, webhooks or reports to the responsible team. RBAC, audit and retention policies govern access to those outputs.