Municipal Utility: AI Assistant for Tariff Questions and Meter Readings

A municipal utility with 120 employees supplies 95,000 households with electricity, gas, and district heating. The customer centre receives around 4,500 calls and 2,000 emails per month. Of these, 58% are standard enquiries: submit meter reading, adjust instalment payment, switch tariff, explain bill, report fault, notify of house move.

Nine customer service staff handled these contacts. During the annual billing season (October–December), contacts exploded by 180% — wait times of 25 minutes were the norm.

Municipal Sensitivity Around Customer Data

A municipal utility is often 100% publicly owned. Passing citizen data — consumption history, payment behaviour, household size — to private US cloud services would be politically and legally indefensible. Full data sovereignty was mandatory.

Local Customer Assistant with SoverIQ

SoverIQ Stack runs in the municipal data centre. The AI assistant is connected to CRM, billing system, and fault ticketing system:

• Submit meter reading and book it directly into the billing system
• Adjust instalment payment on customer request (within defined parameters)
• Calculate tariff comparison and initiate switch
• Explain billing items: “Why is my bill higher this year?”
• Accept fault reports, open ticket, provide expected resolution time
• Record house move and link with new meter reading

Complex cases (final notice, collection, technical edge cases) are escalated to staff.

Results

63% of customer interactions handled fully automatically. Wait times during peak season: from 25 minutes to 2 minutes. Customer satisfaction rose despite — or because of — fewer human interactions. The team now has time for the cases that genuinely require human judgment.