A large motor insurer moved pre-due and grace-window renewal outreach onto governed Voice AI, recovering commitments that used to slip through the cracks, with every call logged and auditable.
Motor renewals are won or lost in a narrow window. A reminder that lands a day late, in the wrong language, or that no human follows up on, becomes a lapsed policy. The insurer had the data to know who was due, but not the reach to act on every case, on time, across a book of millions.
Manual tele-calling covered a fraction of the base, promise-to-pay commitments went unrecorded, and there was no clean audit trail of who was contacted, when, and what was said.
Agentic Universe ran pre-due, on-due and grace-window outreach on a governed AI voice agent. The agent called each customer in their own language, confirmed the vehicle and policy, answered the common questions, and either renewed on the call or sent a tracked payment link to WhatsApp.
Where a customer asked for time, the agent captured a promise-to-pay with a date and set its own follow-up. Nothing was left to a spreadsheet.
This is the difference between a bot that talks and an agent that closes the task. It reads from the insurer's policy and product data, so it never invents a premium or a rule. It handles code-mixed speech, switches language mid-call, and escalates to a human the moment a case is out of policy or high-value.
Every action, from the opening line to the payment link, is logged and exportable for QA and audit.
Outreach ran within calling-hour and consent rules, with a full transcript and outcome on every interaction. High-stakes cases carried human sign-off. Because the agent answers only from verified sources and records everything, the compliance team could review any call in seconds rather than reconstruct it after the fact.
The insurer reached far more of the due base on time, recovered promise-to-pay commitments that previously went untracked, and lifted renewal rate in the grace window, without adding headcount. Handle time dropped and the audit trail moved from partial to complete.
In the words of the operations lead: the work still gets done, every step of it, just without twenty people doing it manually.
Persistency is a connectivity problem before it is a pricing problem. Reaching the right person, on the right channel, at the right time, in their language, and finishing the task on the call, is what moves the number. Governed and auditable is not a feature bolted on afterwards; it is the reason a regulated insurer can run this at scale.
Deployed under NDA and anonymised at the customer's request. Named references available in your segment on request.
A live demo on your own use case, in your language, against your workflow. A real person from our founding team follows up personally.