Managing Collections Across A Large, Distributed Lending Book
Paisalo operates a distributed lending portfolio spanning ABF and MFI businesses, creating a collections environment where borrower engagement must work across large account volumes and different repayment stages.
Fundamento’s deployment focused on automating the outbound layer while keeping field operations connected to the recovery journey.
At Collection Scale, Timing Becomes A Recovery Variable.
Paisalo needed to engage borrowers consistently before and after due dates while changing the intervention as accounts moved deeper into delinquency.
The challenge was to coordinate large-scale outbound activity without reducing every borrower to the same collection script.
| Stage | Operational Objective |
|---|---|
| Pre-Due | Prepare borrower for payment |
| On-Due | Drive timely repayment |
| PD-1 | Early overdue intervention |
| PD-2 | Intensify recovery |
| PD-3 | Prevent deeper roll-forward |
| NPA | Focus recovery on the highest-risk accounts |
Signal OS™ Turns DPD Position Into A Collections Strategy.
The deployment uses DPD position, tenure and engagement signals to determine when and how the borrower should be contacted.
| Signal | What It Revealed |
|---|---|
| DPD / Bucket | Current repayment position |
| Tenure | Borrower lifecycle context |
| Engagement | Response to previous intervention |
| Bucket-Day Position | When outreach should occur |
| NPA Status | Highest recovery priority |
| Architecture Layer | Deployment Capability |
|---|---|
| Customer Signals | DPD • tenure • engagement |
| Decision Intelligence | Bucket-day prioritisation |
| Journey Intelligence | Pre-Due • On-Due • PD-1 • PD-2 • PD-3 • NPA |
| Execution Layer | Voice + IVR + WhatsApp |
| Human Intelligence | Same-day field-team push |
One Collections Engine. Six Repayment States.
| Journey | Objective |
|---|---|
| Pre-Due | Prevent delinquency |
| On-Due | Capture timely repayment |
| PD-1 | Recover early overdue |
| PD-2 | Intensify intervention |
| PD-3 | Prevent NPA progression |
| NPA | Prioritise deep recovery |
| Channel | Role |
|---|---|
| Voice | Primary borrower engagement |
| IVR | Supporting outreach |
| Reminder / payment nudges | |
| Field Team | Same-day physical recovery push |
Multi-Channel Agents Working From One Collections Logic.
| Agent / Layer | Role |
|---|---|
| Voice Agent | Primary bucket-wise collections outreach |
| IVR | Additional borrower reach |
| WhatsApp Agent | Payment reminders and nudges |
| Bucket Logic | Determines treatment by DPD stage |
| Scheduling Logic | Maps outreach to bucket-day grid |
| Field Handoff | Pushes priority cases to field teams |
Automation Covers The Calling Layer. Field Teams Complete The Physical Recovery.
The deployment does not use a human tele-calling floor. Instead, AI manages outbound engagement while priority cases can be pushed to field teams for same-day action.
| AI Handles | Field Teams Handle |
|---|---|
| Pre-due reminders | Physical recovery |
| Due-date outreach | High-priority cases |
| DPD-specific collection calls | Same-day field intervention |
| Payment nudges | On-ground resolution |
Capabilities Are Not Projected. They Are Measured.
38.7%
NPA Share Of Outstanding
100% automated outbound layer
Collections Model
Pre-Due through NPA
Lifecycle Coverage
Voice + IVR + WhatsApp
Channel Coverage
Scale Created A New View Of Collections.
| Intelligence Created | What It Enables |
|---|---|
| DPD Cohorts | Distinct treatment by repayment stage |
| Bucket-Day Intelligence | More structured outreach timing |
| Tenure Signals | Better borrower context |
| Engagement Signals | Identify response patterns |
| NPA Concentration Intelligence | Focus recovery effort where exposure is highest |
| Field Handoff Logic | Connect AI engagement to on-ground recovery |
