Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Digital Services topic
No spam. Unsubscribe anytime.
Finance committee approves four-year AI chat and voice contract with Polymorphic
Summary
The De Pere Finance & Personnel Committee approved a four-year contract with Polymorphic to provide 24/7 AI voice, text and web chat services, citing multilingual access and anticipated efficiency gains; members asked for usage metrics and a year-one usage summary.
Get email alerts on the Digital Services topic
No spam. Unsubscribe anytime.
The De Pere Finance & Personnel Committee voted to approve a four-year agreement with Polymorphic to provide 24-hour AI voice, text and web chat services for residents.
Committee members debated whether the service met an expressed citizen need and how the city would manage accuracy and oversight. Steve, who answered technical questions from the panel, said the system can respond around the clock and reduce staff hours by answering common inquiries: “If you're providing any type of staffing ... after hours ... people can call in 24 hours a day, get answers to common questions,” he said. He added the platform supports multiple languages and produces analytics to track accuracy and usage.
Staff said the service was included in last year's budget as an enhancement, not a replacement for staff, and emphasized an ability to cancel the agreement if it does not meet expectations. Committee members requested a one-year summary showing use, the percentage breakdown of inquiry types and language-service usage; staff agreed the Polymorphic platform includes analytics and that the city will report metrics to the committee.
The committee discussed cost and oversight but concluded the potential to reduce after-hours burdens and improve multilingual access justified approval. The motion passed.
The committee did not set a specific pilot timeline in the meeting; the city will finalize the contract and begin vendor onboarding and training, using department-provided frequently asked questions and website content to train the model.

