The calls that actually fill a utility contact centre — billing questions, outage reports, service starts and stops, payments, appointments — and what an AI agent connected to the CIS does with each of them.
Most of what arrives at a utility contact centre is not complicated. It is repetitive, it is time-sensitive, and it arrives in waves. A handful of call types account for the overwhelming majority of the volume, and they are the same handful at almost every utility we work with.
Here is what Grace does with each of them.
Billing questions
The most common call, and the one most often misdescribed as simple. “Why is my bill higher this month” has a real answer sitting in a rate structure, a budget billing plan, an estimated read that was later trued up, or a longer service period. Grace retrieves the account, explains the actual reason, and sets out the payment options available on that account.
Outage reporting and updates
During an outage the value is almost entirely about capacity. Everyone calls at once, and they call to hear one thing. Grace confirms the location, logs the report, and gives back whatever the utility’s outage system knows — including when the honest answer is that no restoration estimate exists yet.
Service starts, stops and transfers
Move-ins, move-outs and transfers are structured, seasonal, and heavily concentrated at the end of the month. Grace collects what the CIS requires and confirms the service date.
Payments
A failed online payment produces a call within minutes, and it is a call with a deadline attached. Grace can identify what went wrong and take the payment securely, without card details ever passing through the conversation in a way staff or the agent retain.
Account management
Contact details, mailing addresses, communication preferences. Grace verifies the caller before changing anything on the account, and the verification requirement is enforced by the platform rather than left to the conversation.
Usage questions
High-usage calls are billing calls wearing a different hat. Grace gives the consumption breakdown, compares it against previous periods, and explains what changed.
Appointment scheduling
Meter work, inspections, and site visits. Grace finds a slot, books it and confirms it — this is the use case with the clearest measurement, because an appointment is either on the calendar or it is not.
Program enrolment
Budget billing, rebates, efficiency programmes, paperless billing. Enrolment calls are short and they are exactly the calls that get deprioritised when the queue is long.
Emergencies stop the conversation
Gas odours, downed lines, anything with a safety dimension: Grace does not attempt to handle these. She recognises them, tells the caller to hang up and call emergency services or the utility’s emergency line, and routes the call to a person immediately. This is a deliberate limit, not a gap — an emergency is the one call where an AI agent’s job is to get out of the way quickly and unambiguously.
Why this set matters
These call types are the bulk of inbound volume at most utilities, which is why handling them well changes the shape of the whole operation:
- Peaks stop being a staffing problem. Concurrency is what makes a bad day bad, and an agent that takes many calls at once has no queue to manage.
- Staff move to the work that needs them. The calls left over after the routine ones are handled are the complex, sensitive and unhappy ones — the ones a person is genuinely better at.
- Coverage extends past business hours. Nights, weekends and holidays, at the same quality as a Tuesday morning.
Published results from live deployments have run at roughly 60% of interactions resolved end to end without escalation. The remaining 40% still reaches a person — with the context already gathered, rather than starting over.