Conversational AI Speeds Up First Notice of Loss for an Insurance Provider

Industry

InsuranceBFSI

Services

AI Chatbot DevelopmentVoice AI AgentsProcess Automation

Technologies

Azure OpenAIAzure Bot ServiceTwilioPythonReactGuidewire APIAzure

Challenge

An insurance provider's claims contact centre was overwhelmed after storms and other peak events, with policyholders waiting on hold to report claims.

Claims often arrived incomplete, so adjusters spent time chasing missing information before they could assess them:

Long call queues during peak claim periods.
Incomplete claim information causing delays and rework.
Manual data entry into the claims system.
Limited visibility of claim status for policyholders.

The provider wanted a faster, more consistent way to capture claims while keeping a human touch for complex cases.

Solution

Winklix built an omnichannel conversational AI assistant for web, mobile and phone, integrated with the provider's claims platform.

Guided claim intake

Policyholders are verified and then guided through a structured conversation that adapts to the claim type, from motor to property.

Document & photo capture

Customers upload photos and documents in the chat or via an SMS link, and the assistant checks that required evidence is present.

Claims system integration

Completed claims are created automatically in the claims system with a summary for the adjuster, removing manual data entry.

Status updates & escalation

Policyholders can check claim status at any time, and vulnerable customers or complex claims are routed to a human agent.

Insurance Provider solution delivered by Winklix

The Approach

Workflow first — We mapped the real workflow with the people doing the work before choosing models or tools.
Human in the loop — Every AI output had a clear review point, confidence thresholds and an escalation path to a person.
Prototype, then scale — A working pilot on real data proved the value before the solution was rolled out more widely.
Measured from day one — Baseline metrics were captured up front so improvements could be tracked objectively.