Supporting Customers Across A Multi-Product Insurance Environment
Paytm Insurance Broking operates at the intersection of insurance distribution and digital customer service, where customers need timely answers across their insurance-related interactions.
Fundamento was deployed on the inbound side to create an AI-powered first layer for customer service.
Insurance Service Starts With Understanding The Customer's Intent.
Inbound insurance conversations can span multiple service requirements, making intent recognition and appropriate resolution critical to the customer experience.
The challenge was to create a scalable first layer that could understand why the customer was calling and respond or route accordingly.
| Customer Need | Operational Requirement |
|---|---|
| Query Recognition | Understand why the customer called |
| Information Retrieval | Provide the relevant answer |
| Resolution | Complete supported service requests |
| Escalation | Route cases requiring human expertise |
Signal OS™ Starts By Understanding Why The Customer Called.
The inbound deployment uses conversational context to identify customer intent and determine the appropriate servicing path.
| Signal | What It Reveals |
|---|---|
| Caller Context | Who is seeking support |
| Conversation Intent | Why the customer is calling |
| Service Category | Which operational journey is required |
| Resolution State | Whether the request can be completed autonomously |
| Escalation Requirement | Whether human assistance is required |
| Architecture Layer | Deployment Capability |
|---|---|
| Customer Signals | Caller context • conversation intent |
| Decision Intelligence | Service-intent classification |
| Journey Intelligence | Inbound service journeys |
| Execution Layer | AI Voice Agent |
Every Inbound Call Begins With The Right Service Journey.
| Journey | Status |
|---|---|
| Inbound Customer Service | Live |
| Intent Identification | Live |
| AI-led Resolution | Live |
| Human Escalation | Live |
| Intent | Share of engaged | Resolved on call by AI |
|---|---|---|
| General Enquiry | 45.0% | 41.6% |
| Endorsement | 12.7% | 0% (ticket raised for backend) |
| KYC Issue | 12.7% | 48.2% |
| Proposal Issue | 7.8% | 46.3% |
| Cancellation | 7.3% | 0% (ticket raised for backend) |
| Refund | 4.4% | 0% (ticket raised for backend) |
| Policy Documents | 2.8% | 55.6% |
| Inspection | 1.9% | 51.9% |
| Claim Intimation | 1.6% | 0% (ticket raised for backend) |
| Vahan Update | 1.6% | 56.0% |
An AI Voice Layer Built Around Customer Intent.
| Agent / Layer | Role |
|---|---|
| Inbound Voice Agent | Receives and engages customers |
| Intent Intelligence | Identifies service requirement |
| Knowledge Layer | Provides supported information |
| Journey Logic | Determines next service action |
| Routing Layer | Escalates unsupported / complex requests |
AI Resolves What It Can. Humans Take Over Where They Add Value.
| Required Detail | Status |
|---|---|
| Human escalation triggers | Backend action needed (endorsement, cancellation, refund, claims): 37.0% · Explicit request for an agent: 22.2% · Concern not stated: 18.9% · Complaints and escalations: 6.5% · Other: 15.3% |
| Context transferred to human | Pre-filled ticket with the AI call summary, product, request type, language and customer ID |
| Resolution after transfer | 24% |
Capabilities Are Not Projected. They Are Measured.
Live
Inbound Service
Live
AI Voice Deployment
32.7%
Resolution Rate
32.7%
Containment Rate
66 seconds
Average Handling Time
66.1%
Escalation Rate
