Ask the portfolio. Get an answer you can verify.

Ask which building moved off budget, what caused the change and what action is already under way. Every answer links back to the readings and records behind it.

Volts AI assistant answering an office-consumption question with the trend, period and source chart visible.

Ask the business question. Inspect the evidence.

Compare sites, find outliers and explain a change in plain language. The answer names the period and locations it used, then links to the underlying readings so your team can check it.

  • Ask across permitted buildings, utilities and periods in one question.
  • Open the readings, charts and records used to build the response.
  • Continue the investigation without rebuilding the query in another tool.
Volts AI assistant preparing a multi-site consumption report with the source tools and readings exposed.

Keep the answer beside the data it explains.

The assistant appears inside the relevant Volts workflow, already scoped to the site, device or record on screen. You ask in context and move directly from the answer to the next investigation.

  • Pin a conversation to the building, device or workflow currently open.
  • Ask for spend, trend, variance or forecast using ordinary period names.
  • Turn a useful answer into a report or a task.
See dashboards and reports
Volts in-module assistant explaining consumption for the office and device currently open.

The assistant sees only what your role allows.

Answers inherit the same organisation, site and device permissions as the platform. A manager can ask across a portfolio; a local user sees only the locations and records already available to that role.

Answers grounded in
Any site, any period Sources shown Inside every module Your own data
What this means in practice
  • Respect organisation, site and device access on every question.
  • Exclude tenant or financial records the user is not authorised to view.
  • Keep the readings an answer was built from alongside the answer.

Review security and data controls

FAQ

The AI assistant with the limits made clear

What kind of questions can it answer?

Questions supported by the data your organisation has connected: site comparisons, period changes, unusual consumption, missing readings and related operational records. It cannot answer reliably when the required source is absent or outside your permissions.

How do I know an answer is right?

Every answer identifies the sites and period used, and links to the underlying readings or records. Treat the response as a checkable analysis, not an instruction that hides its sources.

Does it see data outside our organisation?

No. It operates within your organisation and role permissions.

How is the assistant different from the agents?

The assistant answers when a person asks. Agents watch configured conditions without being prompted. An agent can surface an exception; the assistant can help a user investigate the evidence before deciding what to do.

Can the assistant change a schedule or control equipment?

Any command from the assistant respects the same site permissions, safety limits and approval rules as manual control.

Can it create a report or maintenance task from an answer?

An answer can become a report or a task without leaving the conversation.

Are our questions and answers stored?

Assistant conversations sit inside your organisation and role permissions, like any other record.

What happens when the assistant cannot support an answer?

It says which reading, period or permission is missing, and stops short of a confident conclusion.

Bring one portfolio question. Check the answer live.

Use a question your team currently answers with exports and spreadsheets. We will show the evidence Volts can reach, the answer it produces and the limits it makes visible.