AI customer assistant for utilities and energy: fewer bill-shock calls, answered from the billing system
How a utility configures Kav: bill-spike and usage questions answered from its own billing system, with a worksheet to find where your money moves.
A large share of a utility's contacts are a handful of questions asked thousands of times: why is this bill higher, what is my usage, when is my payment due. They are cheap to answer when the answer is one lookup away, and expensive when it means an agent opening a billing screen while the customer waits.
A depiction, not a customer. No utility is named here. The example below is the platform's own illustration, marked example data on the home page: the same engine that answers a municipal tax balance, bound instead to a utility's billing system.
What a customer asks, and what Kav answers
- "Why is my bill higher this month?" The assistant answers with both months' totals and both months' usage side by side, taken from the billing system. In the home-page example that is $412.30 against $298.10, 1,120 kWh against 790 kWh. The model chooses which numbers belong on screen; it never writes one. The full bill-spike walk-through shows the detail.
- "Show me usage by day." One tap from the answer, drawn as a real table rather than a paragraph of numbers (why a table is a table).
- "Can I pay this in instalments?" A payment-plan follow-up sits beside the answer. When it needs a person, the agent inherits the whole conversation (what a human agent inherits).
- The channel the customer already uses. The example runs on Telegram; the same capability serves the website and the other channels (one queue, every channel).
Where the money moves
The bold rows are the levers a utility leans on most. Nothing in the last column is a number we supply.
| Lever | What Kav does here | What you measure |
|---|---|---|
| Fewer repeat contacts | Bill-comparison and usage questions are answered from the billing system without a person. | Monthly contacts tagged as bill queries, and the share that are pure lookups. |
| Faster when a person steps in | A payment-plan request arrives with the conversation already attached. | Average handling time on plan requests, before and after. |
| Out of hours, every channel | The same answer at night and on Telegram or the website, not only while the line is open. | Contacts arriving outside opening hours, and what they cost when they wait until morning. |
| Fewer errors and rework | Every figure comes from the billing system through a template, so a reading or a tariff cannot be misquoted. | Corrections and complaints traced to a wrong figure quoted by phone. |
| Revenue kept | A payment plan can be offered at the moment of shock, in the same chat. | Plans taken up after a bill query. |
Work out your own number
- Tag bill-related contacts for four weeks and count how many were a pure lookup.
- Work out your cost per contact: agent hours × loaded hourly cost ÷ contacts handled.
- Monthly value = lookups answered without a person × cost per contact.
Then set it against the plan price. Work out your break-even from the published prices, or talk to us about connecting your billing system.