Understand your fleet, with agents that build its KPIs.
Describe what you want to know in plain language. Specialist agents build the KPIs, analytics, and dashboards; you review and approve; the result runs fleet-wide as live telemetry.
From a plain-language ask to running analytics.
You don't configure analytics screen by screen. You describe what you care about, and specialist agents build it. What you approve becomes part of your asset model itself, so one definition measures every asset of that type you own today or buy next year, with a record of every change. The agents build; a person approves. Nothing runs without your sign-off.
- 01Describe: Say what you want to know, in plain language.
- 02Agents draft: Specialist agents write the calculation, work out which assets it applies to, and build the dashboard.
- 03You approve: Preview every draft over your real history. Nothing activates without an explicit human yes.
- 04Runs as live telemetry: Approved analytics compute continuously on every matching asset as first-class time series. They are charted in the explorer and on asset pages, queryable and exportable like any raw signal, with input quality carried through to every computed point.
KPIs
Larkspur Wind Farm · fleet-scoped
- formula
- available_time / period_time
- inputs
- status_code · fault_code
- window
- 1 day · site timezone
Real-time, without the refresh button.
Dashboards stream straight to the browser, with no refresh and no polling. Per-asset charts and fleet views update as new readings arrive from the edge.
Turbine A07
Larkspur Wind Farm
| Variable | Value | Unit | Quality | Last Updated |
|---|---|---|---|---|
| rotor_speed | 12.45 | rpm | Good | 2s ago |
| active_power | 1240.8 | kW | Good | 1s ago |
| blade_pitch | 27.3 | ° | Uncertain | 5s ago |
| nacelle_temp | 38.6 | °C | Good | 3s ago |
Catch the right signal across the whole fleet.
Six rule types cover the ways industrial equipment goes wrong. Fleet-scoped rules apply one definition across every matching asset, grouping keeps a fleet-wide event from becoming an alert storm, and rules can watch KPI outputs as well as raw signals.
Severity and acknowledge
Alerts carry a severity and move through an acknowledge lifecycle, so an on-call team can triage, claim, and clear incidents without losing track of what is still open.
Webhook and email notifications
Route alerts to the channels your team already uses: outbound webhooks for downstream systems and chat, or email for direct delivery. Scheduled briefings are in early access: a recurring digest on the channels you pick, with a picker that shows exactly who each channel reaches.
Alerts on your KPIs
A rule can watch a KPI output the same way it watches a raw signal. A threshold on round-trip efficiency works exactly like one on a temperature. When corrected history restates a KPI value, the alert follows the restatement, and late data never re-fires a resolved alert.
Explanations with receipts
Early accessAsk why a KPI moved and the assistant answers from the actual inputs. Every claim cites the readings and the definition it came from. Explanations read your data; they never act on it.
Keep exploring
Put your agents to work.
Start with live dashboards and fleet-scoped alerts today, and have your agents build your first KPI in early access.