global-device-insurer

How a Global Device Insurer Saves 10000+ Hours Per Week with AI Agents

A multinational device insurer partnered with StackAI to automate underwriting and claims, cutting turnaround from days to minutes and saving over 10,000 hours per week while improving consistency and customer trust.

A multinational device insurer partnered with StackAI to automate underwriting and claims, cutting turnaround from days to minutes and saving over 10,000 hours per week while improving consistency and customer trust.

Client

Global Device Insurer

Challenge

Manual underwriting and claims reviews consumed thousands of hours weekly. Messy receipts, inconsistent decisions, and delays frustrated staff, slowed policy issuance, and left customers waiting.

Solution

StackAI automated underwriting and claims with OCR, LLM validation, and database checks. Decisions now take minutes, claims resolve instantly, saving 10,000+ hours and improving consistency.

Overview

A leading multinational device insurer operating across multiple countries partnered with StackAI to modernize its underwriting and claims processes. The engagement began after an introduction to the CIO at a dinner, followed by a successful pilot. From there, the insurer expanded into production workflows that now run thousands of times per week.

By deploying StackAI, the insurer replaced two of its most labor-intensive processes, underwriting and claims processing, with automated, AI-driven workflows. The impact has been significant across the organization:

Massive time savings: Tens of thousands of hours of manual review each week have been eliminated. Underwriting that once took days now takes minutes, and claims that dragged on for days are resolved almost instantly.
Improved consistency and compliance: Every policy and claim is validated against the same rules, reducing errors and regulatory risk.
Faster customer experience: Policyholders now receive immediate underwriting decisions and near-real-time claim outcomes, improving satisfaction and trust.
Scalable operations: With AI agents handling the heavy lifting, the insurer can process growing claim volumes without adding staff.
Data readiness: Each workflow outputs structured JSON, making it easy to integrate with databases and analytics systems, enabling further automation.

StackAI now underpins two of the insurer’s most critical workflows, proving that AI-powered agents can deliver measurable efficiency, cost savings, and better customer outcomes at enterprise scale.

Streamlining Underwriting for Device Insurance

The Problem: Manual, Error-Prone Underwriting

Before automation, underwriting new device policies required agents to collect customer and device information by hand. Proof-of-purchase receipts and ID scans were often messy or incomplete, requiring hours of review. Eligibility checks, fraud screening, and risk-tiering were inconsistent across teams, and pricing decisions sometimes varied depending on who performed the review. These inefficiencies slowed down policy issuance, created compliance risks, and introduced friction for both customers and staff.

The Solution: AI-Powered Underwriting Assistant

The insurer piloted StackAI with an underwriting agent designed specifically for device insurance. This agent guides customers and agents through structured data collection, automatically processing uploaded receipts and supporting documents. Using OCR and LLM-powered validation, it authenticates documents, checks eligibility, and assigns a risk tier. From there, it calculates personalized premiums and policy terms, generates a policy summary in Google Docs, and sends an approval email to the customer, all with no human intervention required. The workflow dramatically reduced turnaround time, ensured consistent rules-based decisions, and set a new standard for underwriting efficiency across the organization.

Automating Claims Processing at Scale

The Problem: Unstructured, Messy Claims Intake

Once policies were issued, claims processing remained one of the insurer’s most resource-intensive challenges. Customers submitted claims through email or web forms, often attaching multiple files: phone photos, receipts, handwritten forms, ID copies, and even scanned documents with signatures scribbled across them. Reviewing these by hand took hours, delayed claim resolutions, and introduced significant operational costs. With thousands of claims per week, the insurer’s claims team was under constant strain.

The Solution: End-to-End Claims Processing Agent

After the underwriting success, the insurer expanded with a claims processing agent built on StackAI. This agent ingests all claim documents, applies OCR to extract data from receipts, forms, and photos, and then uses LLMs to structure the data into clean JSON. A first LLM queries the insurer’s Snowflake database to confirm the claimant’s policy details, while a second LLM verifies coverage and evidence against policy rules. The workflow then routes the claim automatically: if approved, it generates a personalized approval email to the customer; if rejected, it sends a clear denial with reasons. Results are logged directly into the insurer’s claim management system. This automation removed nearly all manual review, accelerated claim turnaround times, and gave customers faster, clearer communication.

“When we first looked at their claims intake, it was a jumble of PDFs, phone photos, handwritten forms, and receipts. Adjusters were spending hours just trying to piece together a single claim, and no two submissions ever looked the same. It wasn’t sustainable. By automating the document ingestion and structuring everything into clean data, we gave their team a way to process thousands of claims without drowning in manual review. It’s a night-and-day difference for both the adjusters and the customers waiting on decisions.”

Yiyi Wang
Forward Deployed Engineer

By automating underwriting and claims with StackAI, this global device insurer turned two of its biggest bottlenecks into efficient, scalable workflows. The results speak for themselves: underwriting decisions that once took days now arrive in minutes, claims that dragged on for hours are resolved nearly instantly, and the organization saves well over 10,000 hours of manual work every week. Crucially, this insurer has also achieved greater compliance, consistency, and customer trust, proving that AI agents can deliver real impact at enterprise scale.

Customers