
Playbook
Promise to pay: record it, remind it, check it, and act when it breaks.
A promise written in an agent’s notes is not a promise anyone will follow up. Most broken promises are simply never chased.

A promise written in an agent’s notes is not a promise anyone will follow up. Most broken promises are simply never chased.
A borrower says they will pay on the tenth. The agent types it into a note. The tenth passes, nobody checks the payment file against the note, and the account rolls into the next bucket.
| Metric | Definition |
|---|---|
| Promise rate | Promises taken ÷ borrowers contacted |
| Promise kept rate | Promises paid in full by the date ÷ promises due |
| Re-queue time | Time from broken promise to next contact |
| Repeat breakers | Accounts with more than one broken promise in the cycle |
AutoCollect is built for collections teams at banks and NBFCs. Due and overdue accounts are loaded daily from your LMS and ranked every morning by propensity to pay, with the best channel for each borrower. Reminders and payment links go out on WhatsApp before and after the due date; voice agents take promises in the borrower’s language; disputes, hardship and restructuring always go to a trained person; later buckets are routed to field agents with proof of visit. Payments are matched to promises and broken promises re-queued automatically. Contact-hour rules, approved scripts and a do-not-call list apply on every channel, and every call, by people and voice agents alike, is summarised and scored. Your policy decides what can be offered.
A borrower’s commitment to pay a specific amount on a specific date, recorded so it can be reminded and checked against payments.
The account should be re-queued quickly, often to a different channel or a person, and repeated breaks should change the treatment under your policy.
See it on your own data. AutoCollect — The right borrower, the right channel. Book a 30-minute working session with an engineer.