Purestream · A water network which sees failure coming

Purestream moved produced water for operators across the Permian basin. The volumes dwarf the oil they travel with, and the margin rides on uptime. The network ran on radio calls, spreadsheets, and drive-the-line inspections; the data existed, but it had no shape.

Crews chased incidents after the fact. Executives saw which stations had failed, never which were about to. Every blind day had a price, in barrels, in trucks, and in trust.

Six weeks embedded with operations, field leads, and executive sponsors. We rode the routes, sat control room shifts, and logged where information stalled. Two decisions carried the whole operation: which asset gets attention first, and how much loss is on the table today. Everything else would serve those two answers.

The stations were instrumented. The trouble sat between them: pressure fading between pump and line, custody blurring between intake and reinjection, alerts landing in channels the field never read. The finding redirected the build.

The bet: one live map which carries the whole network, intake to reinjection, and prices every deviation in dollars per day. Not another dashboard beside the others. The one surface every role reads, from the control room to the truck. If it is not on the map, it does not exist.

We defined the experience as a promise. No operator learns about their own network from a phone call. A pressure drop scores itself, prices itself, and reaches the nearest crew as a task with guidance in hand.

The same picture at every scale, in every hand. Color is the information. An alert lands as a place, not a row in a table.

The build assembled Intelligent Core engines against the map. Ingestion meets the field as it is: SCADA, distributed IoT telemetry, and audit-ready chain-of-custody logging, drawn into one control layer. The platform was engineered to process millions of telemetry data points across distributed water assets in real time. Insight scores the risk. Foresight prices the loss before it compounds. Relay turns a prediction into a dispatched crew, and the field tablet carries the same picture as the control room.

Built end to end by the studio: strategy, product, brand, and experience. One team, one voice, briefing to launch.

Team. Robert Girvin and Tim Rowe, founders. Dr. Ndidi Nzeako Anyakora, data science and machine learning.

Client. Purestream, then led by CEO Adam Gilles. The outcomes above were measured in his operation.

Discipline. Industrial water · environmental technology · predictive intelligence.

Ownership. Studio-built, client-operated.