How we deliver
No data is better than bad data. Every data point we deliver is validated, labelled, and monitored so you can act on it with confidence, whatever you use it for.
Book a demo
Every incoming data point goes through 8 validation checks: utility type, unit, timezone, flow, measurement type, rate calendar, aggregation level, and invoice match. Data that fails a check is flagged before it enters your dataset.
Statistical outliers, negative consumption, decreasing readouts, and future timestamps are flagged automatically. Raw data is never silently altered. Anomalies are surfaced so they can be investigated and resolved at source.
Every meter is monitored continuously. A 7-day buffer applies to daily meters; 30 days for monthly. When a connection is interrupted, we detect it and act before it becomes a gap.
We track the percentage of expected timestamps with valid data, per meter. Where gaps exist, we attempt recovery from the provider. Where that isn't possible, invoice data fills the remaining gaps.



