Skip to content

AI Product

Pilot

Water AI

Water management for utilities, industry and agriculture.

In one sentence

Water networks lose a large share of what they pump, and demand swings with weather and season. Planning still runs on last year's numbers and leaks are found when someone reports a wet street.

The problem

Water is managed on averages. A utility plans on last year’s demand, a plant on its permit, a farm on the calendar. Meanwhile the network leaks, a heatwave doubles demand in a district, and the first sign of a burst is a phone call. The data to do better often exists, in flow meters and SCADA logs, but it sits unused.

How it works

Water AI connects to the sensors and systems a water operator already has and turns the readings into three things: a demand forecast per zone that planners can act on, leak and burst signals from night-flow and pressure patterns, and dashboards that show where the water actually goes. For agriculture, the same models drive irrigation schedules from soil, weather and crop data.

It builds on our forecasting work with the Athens water utility and on open research the founder has published.

Where it is today

Water AI is in pilot. We are working with a small number of utilities and agricultural partners to validate the models on their networks. If you manage water at scale and want to be part of the next pilot, get in touch.

What it does

Features that matter.

Demand forecasting per zone

Consumption forecasts by district, day and hour, aware of weather, season and holidays.

Leak and burst signals

Night-flow and pressure patterns that point to leaks before they surface.

Irrigation scheduling

Water only when and where crops need it, from soil, weather and crop data.

Consumption dashboards

Where the water goes, by zone, customer group and time, in one view.

Alerts for operators

Threshold and anomaly alerts routed to the people who can act.

Sensor ingestion

Flow meters, pressure sensors and SCADA feeds in, with gap handling built in.

Interested in Water AI?

We are onboarding a small number of partners. Tell us about your setup and we will get back within one business day.