For traders · Operators · Portfolio managers · Aggregators
Customer-tailored solar generation forecasts built with proprietary deep-learning models for assets, portfolios and electricity markets.
GraphSight uses Graph Neural Networks to process spatial relationships, new weather information and operational signals together. The result is a forecast shaped around your data and decisions.
Three decision levels
Forecast a single plant, the portfolio you manage, or the market around you.
Your plants
Use Asset Forecasting when the performance of the individual plant matters. Each site is modelled independently using its location, technical characteristics and operational conditions.
Best for: Asset managers, operators and trading teams managing plant-level positions.
Explore Asset ForecastingYour position
Use Portfolio Forecasting when multiple solar assets are managed as one commercial position. The portfolio is modelled jointly to capture diversification, changing composition and shared market exposure.
Best for: Utilities, aggregators, IPPs and trading desks managing multiple assets.
Explore Portfolio ForecastingThe market
Use Zonal Forecasting to estimate total solar production across a bidding zone or control area for trading, fundamental analysis and intraday market decisions.
Best for: Power traders, analysts and market participants looking beyond their own portfolio.
Explore Zonal Forecasting| Decision level | AssetYour plants | PortfolioYour position | ZonalThe market |
|---|---|---|---|
| Forecasts | One or more plants | Your combined portfolio | The wider market |
| Question answered | What will these plants generate? | What position will I trade? | What will the market generate? |
| Primary use | Operations & nominations | Portfolio trading & imbalance | Market analysis & trading |
| Level | Sites | Commercial portfolio | Bidding zone |
Why GraphSight
Forecast quality matters, but it is only useful when the service can integrate with your systems, scale across real portfolios and react when new information arrives.
Use models designed for the selected scale, horizon and operating context, evaluated against the data and reference that matter to you.
Handle large sets of assets and deliver farm-level forecasts without turning every additional plant into a separate integration project.
React when fresh information becomes available, including new NWP runs, satellite imagery and operational signals.
Exchange data through the interface and format that fits your workflow, from REST APIs to file transfer over S3 or SFTP.
Operations-aware forecasting
Irradiance and cloud movement shape solar potential. Operational signals help forecast what will actually reach the grid.
See the technologyDistinguish grid- and market-driven constraints from weather-driven production.
Model suppressed production and recovery as a separate physical effect.
Reflect the plant or portfolio that is actually available.
Service tiers
Select Asset, Portfolio or Zonal coverage, then choose how closely the service should fit your operating context.
A clearly defined forecasting scope and delivery setup.
View tierAdapted to your data, geographic exposure, forecast horizons and evaluation criteria.
View tierTailored zonal forecasts with handling for specified operational conditions.
View tierTell us the coverage, horizons and level of adaptation you need, and we’ll define the right setup.
Forecast performance depends on geography, plant behaviour, weather regime, lead time and data quality. Test GraphSight against the forecast and actual production data you use today.
Define assets, period, horizon, issue times, metrics and data-quality rules.
GraphSight sends you the forecasts following the agreed specs and format.
Compare GraphSight with your current forecasts before deciding how to proceed.
Insights
Tell us the assets, portfolio or zone you need to forecast. We will review the available data and propose a fair evaluation scope.
