eSecuritel

The Challenge

Mobile device insurance in its traditional form was a customer experience failure. Policyholders filed claims through manual processes — phone calls, paper forms, mailed documentation — that took days to resolve and produced outcomes that felt arbitrary. Fraud was assessed subjectively or not assessed at all, creating loss ratios that made the product economically fragile at scale. Customer satisfaction was predictably low.

The underlying problem was that mobile device insurance had scaled as a distribution product without scaling as an operational product. Carriers could sell policies at volume. They couldn't process claims at volume without proportional headcount growth, and proportional headcount growth destroyed unit economics.

The brief: rebuild the claims operations layer using computer vision and machine learning to automate damage assessment, accelerate resolution times, and detect fraud at a level that the manual process could never achieve — while delivering a customer experience that made policyholders feel served rather than processed.

---

The Approach

### Computer Vision for Damage Assessment

The core innovation in eSecuritel's claims pipeline was the application of computer vision to physical device damage assessment. When a policyholder filed a claim, the platform guided them through a standardized photo capture workflow — front, back, ports, screen — and the computer vision layer evaluated the images for damage type, severity, and consistency with the claimed incident.

This replaced the most time-consuming and subjective part of traditional claims handling: the manual review by an adjuster who would evaluate photos, request additional documentation, apply inconsistent judgment, and create resolution delays measured in days. The CV pipeline assessed damage in seconds and produced a structured output that drove the resolution decision automatically for straightforward claims.

The Results

The Outcome

Brightstar — one of the world's largest mobile device lifecycle management companies, since rebranded as Likewize — acquired eSecuritel because LumenIQ had built something that strategic buyers pay to own: a defensible operational capability at scale. The 500,000 subscriber base was valuable. The 95% fraud detection accuracy was valuable. But what made eSecuritel acquisition-worthy was that the technology worked, the unit economics held, and the customer satisfaction score was 92% in a category where customers expect to feel cheated.

The eSecuritel exit established LumenIQ's pattern as an AI product builder: identify a process that humans are doing at scale and doing poorly, build the AI infrastructure that does it better, scale to the point where the economic value is undeniable, and build with the quality that survives the scrutiny of a strategic acquirer's due diligence.

That pattern has been applied to every product in the portfolio since.

---