Recover more, earlier, inside the rules.
AI agents engage borrowers across voice, WhatsApp, SMS and email — based on DPD, history and channel preference — inside every RBI recovery conduct requirement.
What breaks without governed AI
Contact rates fall as buckets age
By 30+ DPD, many borrowers stop answering calls. Without channel intelligence, teams escalate to field too early and drive up cost-per-rupee-recovered.
Single-channel calling misses borrowers
Each borrower has a channel and time they respond to. A script that calls at the same time to the same number regardless of history produces low contact rates and borrower complaints.
Conduct breaches found after the damage
Calls at prohibited hours, agents using pressure tactics, no record of consent — recovery conduct violations are discovered by auditors after complaints reach the regulator.
What the agent does, step by step
Every action is governed, logged and traceable. Your team stays in control at every checkpoint.
The agent connects to your LMS to understand which bucket the borrower is in, their payment pattern, and what channel and time they have previously responded to.
Based on history and your configured rules, the agent selects WhatsApp, SMS, email or initiates an outbound call — only within RBI-permitted hours for that borrower.
For eligible accounts, the agent presents EMI deferment, partial payment or a settlement option — only from the menu you have pre-approved for that product.
Every outcome — delivered, read, promised, refused, or no response — is logged with a timestamp and written back to your collections system within seconds.
Distressed borrowers, disputes, or anyone who explicitly asks for a person are escalated immediately. Your agent gets the full conversation log, not a summary.
“A call at 7:40 pm never happens.”
Calling hours, contact frequency, permissible messages and approved settlement ranges are configured once — and enforced on every contact, every time, with no manual check required.
The measures every pilot is judged on
We agree the scorecard before the pilot begins. No metrics are invented after the fact.
Contact rate by DPD bucket
Measured vs. baseline in your pilot
Promise-to-pay kept rate
Measured at 7 and 30 days post-PTP
Roll-rate between DPD buckets
Movement from 30+ to 60+ DPD vs baseline
Complaints per 10,000 contacts
Reported to your compliance function
Cost per rupee recovered
Measured across channel and bucket
Field escalation rate
Cases reaching field vs. resolved remotely
Sits on your stack. Reads and writes back.
Vaman connects to the systems you already run — reading context from them and writing every outcome back. No new data warehouse, no manual reconciliation.
See the full integration mapWhat people ask before they start
Recover more. Without conduct risk.
A readiness assessment gives you a prioritised use-case plan, a data readiness report and a pilot scope your board can approve — in three weeks, for a fixed fee.