Voicana

AI speech analytics for detecting fraud during insurance claims calls.

Claims Technology Startup

Overview

Voicana is a claims fraud detection platform for insurance carriers, using real-time speech analytics on phone-based claims calls.

The system analyzes vocal characteristics such as tone, pitch, intonation, and rhythm alongside the semantic content of a call, combined with emotional-marker detection for stress, hesitation, and anger, to generate a live risk score for the claims handler during the call. Machine learning models are continuously retrained on a library of historical call recordings. The platform is built to handle large call volumes on dedicated servers and integrates with a carrier's CRM or ERP systems for data exchange.

No named customer deployments or disclosed funding rounds were found. The product is developed by Commint Ltd (trading as Commint.pl), based in Nowa Iwiczna, Poland, near Warsaw.

Products & Services

Voicana AI Speech Analytics

Real-time analysis of insurance claims phone calls that flags fraud risk as the call happens, combining prosodic analysis of the caller's voice with semantic analysis of what is said.

Key Features

Combines tone, pitch, intonation, and rhythm analysis with semantic/linguistic analysis of the spoken content
Detects emotional markers such as stress, hesitation, and anger
Produces a live, per-call risk score visible to the claims handler during the call
Machine learning models continuously retrained on a library of historical call recordings

Target Users: Claims handlers at insurance carriers