Capabilities · AI

Every meter event is training data the moment it lands.

JSON in, signal out. Loss detection, demand forecasting, credit underwriting, fraud flagging — built on data the meter generates as a byproduct of doing its job.

What the data trains

Loss detection. Tamper, theft, meter bypass, technical loss — pattern-matched across the fleet.

Demand forecasting. Per feeder, per substation, per hour, per season.

Credit underwriting. Two years of payment behaviour. Underwritable. Portable. Consented.

Fraud flagging. Anomalies in payment, load, location — surfaced before the bill clears.

Why it matters

Every other meter on the market generates data the vendor sells back to the operator as a feature. We generate data the operator owns from the moment it leaves the device.

The model trained on that data is yours.