Two operating batteries. Two fitted decision programs.

We rebuilt real ERCOT operating histories, modeled each battery, and replayed what Metis would have done under the same market and physical boundary.

The public record lets us reconstruct the decision path, fit an asset-specific program, and put every candidate through the same promotion boundary it would face before live operation.

Same battery. Same days. More margin.

A replay of 107 operating days found $585k of additional energy margin. Roughly $55 per MW-day on a 100 MW asset. It held after we deleted the ten best days.

The asset's own dispatch$909k
Metis replay candidate$1.493m
After-variable-cost energy margin · central sensitivity

+$585k

additional after-variable-cost energy margin across 107 operating days — about $55 per MW-day on a 100 MW / 200 MWh asset

Public-data replay

Every sensitivity in the registered envelope came back positive, from +$429k to +$619k.

Read the operating study →

Fitted to one battery. Challenged before deployment.

We rebuilt 121 days of operating history, modeled the battery, and fitted a receding-horizon policy to its physical and market context.

operating history reconstructed
121 days
decision cadence
15 min
receding planning horizon
24 hr

121 days

of operating history replayed at 15-minute decision cadence with a 24-hour receding horizon

Asset-fitted replay

A public physical twin, forecast fan, and receding-horizon policy were fitted to one battery's operating record.

Read the operating study →

Now fit the same program to your operating truth.

Benchmark an asset →