SolarPredict AI
SolarPredict turns site inputs into a right-sized system design, a location-aware generation forecast and actual-vs-predicted SGF analytics — replacing slow, manual sizing spreadsheets.
Built to run the work, not just answer questions.
System sizing
Recommends capacity, module count and inverter from load, roof area and location — with tilt & azimuth capture and editable losses.
Generation forecast
Predicts expected output from PVGIS / NASA POWER / Open-Meteo irradiation through the plant Performance Ratio, so proposals are grounded in data.
Actual vs prediction
Tracks SGF (generation ÷ capacity), deviation %, MAPE and revenue loss — with month-level alerts and underperformance ranking.
See it working on real scenarios.
A live, gated portal on an API-first backend. Sign in with a demo role to explore SolarPredict AI against representative, non-confidential data.
- 8-role RBAC, configurable by Super Admin from the UI.
- SGF actual-vs-predicted with revenue-loss analysis.
- Grounded AI assist over your database and knowledge base.
Secure by design
Role-based access, audited actions and demo data only — no confidential customer information is exposed in this environment.
- Sign-in required — data persisted via backend APIs, not the browser.
- Every add / edit / export logged for audit.
- Customisable to your own workflow and branding.