Smart EMI Collections with Dynamic Tone

A digital lending team replaced generic payment reminders with a WorqHat workflow that tailors the tone of every EMI message based on days overdue.

Business impact:

Polite nudges for fresh delays
Firm reminders for mid-cycle delays
Legal-tone escalation for chronic defaults
Higher collection rates with better customer experience

The entire pipeline runs automatically every morning.

Previous State

Support team manually scanned overdue loans →

Checked how late each payment was →

Copy-pasted different template messages →

Sent WhatsApp reminders individually

Result:

Slow. Inconsistent. Zero personalization. Lower recovery rates.

Target State

A 7-node automated WorqHat workflow handles the entire collection cycle:

Calculates delay for each customer
Classifies tone based on severity
Generates the correct message tone via AI
Sends the WhatsApp message instantly
Runs daily with no agents involved.

Workflow Breakdown

1. Time Based Runs (Trigger)

Purpose: Start the collection routine every day at 10 AM.

Output: Initiates the pipeline.

2. Query Data (Fetch Overdue Loans)

Query:

payment_status == "Overdue"

Output: Array of overdue loans

(loan amount, customer name, phone, due_date, etc.)

3. For Loop (Process Each Loan Individually)

Iterates through each overdue record retrieved.

Provides loop context for loan-specific processing.

4. Custom Node (Calculate Days Overdue)

Logic:

days_overdue = Today – due_date

Output: Integer days late

5. Dynamic Tone Logic

Instead of multiple branches, this workflow uses one AI prompt that adapts tone internally based on days_overdue.

Tone categories:

1–7 days: Polite
8–30 days: Firm
30+ days: Stern / Legal

6. Text Generation (Draft Message)

Prompt:

Write a personalized WhatsApp reminder for {{name}} about the overdue payment of {{amount}}.

Payment is {{days_overdue}} days late.

Tone rules:

- 1–7 days → polite
- 8–30 days → firm
- 30+ days → urgent, mention credit score impact.

Output: Final WhatsApp message text.

7. Send WhatsApp (Deliver Reminder)

Input:

phoneNumber → borrower's phone
message → AI-generated content

Sends the message instantly.

8. Return State

Output:

{
  "status": "success",
  "message": "Collection run completed",
  "data": "total overdue accounts processed"
}

Outcomes

Higher repayment rates through tone-matched messaging
Zero manual chasing for support teams
Consistency at scale — new defaults handled instantly
Improved customer experience with contextual communication
Automated escalation without rewriting templates

Recommended Extensions

Add auto-escalation notifications to internal collections team
Add payment link generation per message
Add multi-language AI messaging (English/Hinglish/Regional)
Add AI categorization of chronic defaulters

Next Steps

Use this pattern to automate any periodic behavioral messaging:

Invoice reminders
Subscription renewals
Maintenance alerts
Policy expiries

Simple rule:

If tone needs to change with time — automate it.

👉 Install this template in WorqHat and upgrade your collections workflow.

Launch "Smart EMI Collections with Dynamic Tone" as a workflow