Quantified outcomes
by industry

The Challenge

Meridian was losing customers without warning. By the time churn showed up in quarterly attrition reports, the customer had usually already decided to leave.

Separately, account takeover fraud was increasingly difficult to catch using rules-based detection alone, since fraudulent sessions often looked credential-valid on the surface.

The Solution

OntarioAI deployed Pulse to analyze sentiment across customer service interactions, surfacing early churn risk signals weeks before account closure.

For fraud specifically, Pulse also analyzed behavioral biometric signals, typing and interaction patterns, to flag likely account takeover attempts. Nudge routed churn signals to the retention team and fraud signals to the fraud team in real time, both integrated with Meridian's existing CRM and fraud systems in under six weeks.

The Challenge

Northstar's fraud and quality team could only manually review a small sample of claims calls each month. Fraud indicators in calls outside that sample went undetected until much later in the claims lifecycle, if at all, and call quality auditing suffered from the same coverage gap.

The Solution

OntarioAI deployed Pulse to analyze claims call sentiment and language patterns across the full volume of calls, not a sampled subset, supporting both call quality auditing and fraud indicator detection during claims handling.

For claims flagged as high-risk, Pulse also analyzed behavioral biometric signals to strengthen fraud mitigation accuracy. Nudge routed flagged claims directly to the fraud investigation team in real time.