Everything from what we forecast to how scenarios, uncertainty, delivery, and security work—so you can decide if Full Load Forecasting fits your workflow.
All-inclusive price and load forecasts at node and zone levels, trained on hundreds of inputs. You can adjust assumptions (e.g., fuel spreads, outages, renewable bias) and immediately see the impact on both inputs and predictions. Forecasts ship with confidence intervals, variance, rate-of-change, and probability-of-exceedance to make risk explicit.
We focus on North American ISOs/RTOs at launch (e.g., PJM, ERCOT, CAISO, NYISO, MISO, ISO-NE, SPP). Support is expanding—if you have specific coverage needs, contact us.
No. We ingest and maintain the core exogenous signals (ISO feeds, weather, outages, fuel spreads, policy calendars) and map your nodes/zones during onboarding. You can optionally add private feeds later; those remain isolated to your project.
Use scenario controls to bias or shock key drivers—renewable output, outage likelihood, demand multipliers, constraint toggles, fuel spreads—and watch both feature inputs and forecasts respond in real time. Changes are non-destructive and fully traceable, ideal for what-ifs and reviews.
Yes. Drag-and-drop projects onto the map, set sizing and parameters, and the model adapts local prices using topology-aware features and constraint stress. Compare before/after spreads, volatility, and exceedance probabilities to support siting and timing decisions.
Forecasts cover near-term through medium-term horizons typical to trading and planning workflows. Granularity matches market cadence (e.g., day-ahead/real-time) with roll-forward updates and online calibration for regime changes.
Choose 50/80/95 (and more) to widen/tighten bands. We expose variance, rate-of-change, and probability-of-exceedance so you see the distribution—not just a point estimate. Calibration is monitored and corrected over time.
Regime gating plus online calibration maintains stability across weather, fuel, and behavioral changes—avoiding brittle full retrains. Diagnostics are segmented by regime for honest performance readouts.
Latency-correct out-of-sample evaluation (simulate data arrival), strict leakage guardrails, and reproducible baselines. Backtests are exportable, and we prefer comparisons against frozen benchmarks.
Standard HTTPS API for reads; SSO/MFA on the hosted app; optional secure file drops for scheduled exports. If you have a specific integration path, tell us and we’ll align on the cleanest approach.
We host on AWS with industry-standard controls. We minimize collection of personal data, isolate customer-specific inputs, and maintain least-privilege access for operations. See our Privacy page for data practices.
We map your nodes/zones, confirm coverage, and configure access. Typical setup is measured in days, not months, depending on number of users and delivery paths (dashboards, API, alerts).
Many teams run us alongside internal models. Our value is adaptive performance, transparent diagnostics, and operational speed. Use our forecasts as a primary signal or as a check on your stack—your choice.
We scope access by team size and delivery needs (dashboards, API, alerts). A focused pilot is often the fastest way to evaluate value on your nodes. Contact us and we’ll propose a fit.
Yes. The hosted app supports SSO/MFA. See Login for access paths (demo vs. client).
Email support for all customers, with options for shared Slack channels and escalation SLAs on commercial plans. We also provide review sessions for scenario design and validation when helpful.
Outputs are for informational purposes and are not investment advice or an offer to trade. Please review your internal policies and our Privacy terms for data handling.
Tell us your nodes, timelines, and goals—we’ll respond with a short plan and a focused pilot option.
Contact us
Full Load Forecasting