Demand forecasts built for power traders, grid operators, and utility teams.

Use Gramm when day-ahead and multi-day load forecasts affect bids, hedges, dispatch, staffing, or risk. Review the benchmark, test side by side, then switch when the numbers hold.

See the same things your desk, engineers, and procurement team will ask for.

Public benchmark tables by market, with the baseline visible.

Reports, methodology notes, and API docs before a sales call.

Status, SLA, security, and procurement pages for formal review.

Used for Day-ahead bidding, weather-driven trading, load planning, dispatch, and storage operations.

California. Current public benchmark moves day-ahead MAPE from 4.8% to 2.9%.

Coverage. Day-ahead, 48-hour, 7-day, and 15-day load forecasts.

Formats. OASIS-style requests with JSON, CSV, and XML responses.

Operating use. Built for bidding, hedging, dispatch, and load planning.

Switch support. Backtests, status, SLA, security, and procurement review.

Start with the markets where load errors get expensive.

The strongest case for switching is not a slogan. It is a market, a benchmark, and a clear explanation of why the difference matters operationally.

39.6% lower day-ahead error in the current California public benchmark

4.8% -> 2.9% MAPEpublic benchmark

For traders and operators, tomorrow's load forecast shapes bids, hedges, staffing, and unit plans. The current California comparison is the clearest public view of how Gramm performs against the published baseline.

Read Benchmark Memo

Tail misses matter more than average days

60% fewer large misses

In ERCOT, the expensive days are the ones that matter. This is where weather turns, scarcity risk, and large forecast misses hit hardest for traders and operators.

Read Monthly Report

One feed for tomorrow morning and the rest of the week

Day-ahead to 15-day

Trading desks and operating teams do not want one vendor for short-horizon load and another for planning. Gramm covers both on the same API.

Review API Docs

The public benchmark is the first thing to check.

Review territory-level results with the baseline, error metric, and reduction shown in plain view. If the public table is interesting, the next step is a side-by-side backtest for your market and horizon.

Territory
Baseline
Gramm
Reduction
California
CAISO · PG&E, SCE, SDG&E
4.8%
2.9%
39.6%
Texas
ERCOT · Oncor, CenterPoint, AEP Texas
5.1%
3.4%
33.3%
Mid-Atlantic & Midwest
PJM · PECO, ComEd, Dominion
3.9%
2.6%
33.3%
Central U.S.
MISO · Entergy, Xcel, DTE
4.3%
2.9%
32.6%
New York
NYISO · Con Edison, National Grid
4.1%
2.8%
31.7%

The public table is a starting point. Buyer-specific backtests add the dates, sample size, issuance timing, and horizon detail for your market.

Traders, operators, and flexible load teams switch for different reasons.

The common requirement is the same: the forecast has to improve real decisions, not just look good in a model chart.

Power traders

Better load forecasts matter when bids, hedges, and weather risk are time-sensitive. That is especially true on the days when demand and price move together.

Grid and utility operations

Lower error helps with staffing, generation scheduling, reserve planning, and next-day operating decisions.

Storage and flexible load

Dispatch improves when the load view is timely, repeatable, and available across both short and medium horizons.

Benchmark before the call

The public benchmark gives buyers something concrete to check before they spend time with sales.

Easy cutover

OASIS-style requests and structured outputs reduce integration work for engineering teams.

One vendor across horizons

No handoff between trial, live short-horizon use, and 7-day or 15-day planning.

A standard evaluation has four steps.

1. Check the public benchmark

See how Gramm compares with the published baseline in the markets you care about.

2. Run a side-by-side test

Request a backtest for your desk, utility, or operating region and compare it with the feed you use today.

3. Confirm delivery

Check API behavior, formats, auth, and update timing with engineering before cutover.

4. Cut over with support

Move live when trading, operations, and procurement are comfortable with the switch.

Tell us the market, horizon, and use case.

We will send a backtest package sized for your desk, region, or operating team, plus the technical and commercial documents needed for review.