01
Bill & data intake
Structure billing periods, usage, fixed charges, delivery charges, and available interval records.
Platform / Utility intelligenceEvery number needs
Every number needs
the right context.
Utility AI connects building records, regional pricing, weather, and operating events—then turns the strongest evidence into a useful decision brief.
Analysis depth follows the quality and availability of the source data.
Context in
Bills & usage01
Rates & location02
Weather & systems03
Utility AIEvidence engineDecisions out
Cost drivers
Signals to verify
Ranked next steps
Core intelligenceFrom raw records
From raw records
to practical findings.
Start with recent bills and a basic building profile. Add deeper usage and operating information only when it improves the decision.
02
Regional rate context
Apply supported provider, municipal, and time-based pricing information to the right periods.
03
Weather-informed review
Compare heating and cooling changes with seasonal and temperature conditions.
04
Possible anomaly signals
Surface unusual patterns that may deserve management or maintenance verification.
05
Cost & usage forecasts
Estimate what may come next from available history, rates, weather, and known changes.
06
Decision-ready reporting
Present findings, evidence, assumptions, confidence, and practical next steps clearly.
Utility modulesOne platform.
One platform.
Three analytical lenses.
Each utility has different cost drivers, patterns, and data constraints. Those differences stay explicit.
Sample data
01 / Electricity
Rates, peaks, and load
Understand how usage timing, pricing structure, and building loads influence the bill.
- Supported TOU, ULO, tiered, or tariff context
- Peak and high-cost period review
- Flexible-load opportunities
- EV, electric heating, or cooling context
Peak 126 kW at 17:00. Cost also depends on the rate plan.
| Hour beginning | Mean demand kW |
|---|---|
| 00:00 | 38 |
| 01:00 | 35 |
| 02:00 | 33 |
| 03:00 | 32 |
| 04:00 | 34 |
| 05:00 | 40 |
| 06:00 | 55 |
| 07:00 | 68 |
| 08:00 | 72 |
| 09:00 | 76 |
| 10:00 | 81 |
| 11:00 | 85 |
| 12:00 | 89 |
| 13:00 | 94 |
| 14:00 | 101 |
| 15:00 | 110 |
| 16:00 | 118 |
| 17:00 | 126 |
| 18:00 | 112 |
| 19:00 | 93 |
| 20:00 | 72 |
| 21:00 | 60 |
| 22:00 | 49 |
| 23:00 | 42 |
02 / Water
Patterns and baselines
Review monthly and overnight behaviour beside building history and occupancy context.
- Municipal billing information
- Possible abnormal-use indicators
- Occupancy-normalized comparison
- Unit-level support where authorized
Dashed: 1,087.9 m³ mean. Check billing periods and occupancy.
| Month | Water m³ |
|---|---|
| Jul 2025 | 1020 |
| Aug 2025 | 1015 |
| Sep 2025 | 990 |
| Oct 2025 | 1005 |
| Nov 2025 | 995 |
| Dec 2025 | 1010 |
| Jan 2026 | 1080 |
| Feb 2026 | 1120 |
| Mar 2026 | 1160 |
| Apr 2026 | 1200 |
| May 2026 | 1220 |
| Jun 2026 | 1240 |
03 / Gas & heating
Season and performance
Compare heating demand with temperature, building systems, and known operating events.
- Heating-season consumption
- Weather-informed usage review
- Boiler and system context
- Next-period cost forecast
Dashed: weather estimate. Shaded: comparison band ±10%.
| Month | Observed gas m³ | Heating degree days (example) | Model estimate m³ |
|---|---|---|---|
| Jan 2026 | 17000 | 625 | 16700 |
| Feb 2026 | 15600 | 565 | 15500 |
| Mar 2026 | 12100 | 385 | 11900 |
| Apr 2026 | 8100 | 200 | 8200 |
| May 2026 | 5900 | 80 | 5800 |
| Jun 2026 | 4800 | 30 | 4800 |
Context stack
A finding is only as useful as the evidence beneath it.
Utility AI keeps each source layer visible so teams can understand what shaped the finding—and where more information would improve confidence.
No hidden certainty. Limitations and assumptions remain part of the output.
Utility record
Bills, usage, charges, readings, and available interval files.
Regional context
Location, providers, municipalities, supported rate plans, and effective dates.
External conditions
Temperature, seasonality, heating and cooling demand, and forecast conditions.
Building profile
Units, floor area, age, occupancy, systems, amenities, and recent changes.
Questions worth answeringMove from “what changed?”
Move from “what changed?”
to “what should we check?”
The platform is designed to investigate practical building questions without turning an indicator into an unsupported conclusion.
Did usage, timing, pricing, or a new load drive the increase?
Is the rise consistent with occupancy—or worth a leak investigation?
Was the change weather-driven, or unusual after normalization?
What might the next period cost under current conditions?
Designed for responsible review
Useful intelligence. Explicit boundaries.
Signal, not diagnosisPossible anomalies are items for verification, not claims of failure.
Evidence, not opacitySources and assumptions stay connected to each finding.
Confidence, not false precisionThe output reflects the quality and completeness of available data.
Decision, not decorationEvery priority signal leads to a useful next review step.
Complimentary pilot
Request pilot See what one building can reveal.
Start with a focused pilot and the utility question that matters most to your team.