Intelligent Claims Handling: How Van Ameyde is Transforming Claims with AI

Introduction

On May 28, 2025, Van Ameyde Group hosted a forward-looking webinar titled “Intelligent Claims Handling”, highlighting the company’s innovative approach to streamlining the claims process through automation and artificial intelligence (AI). The session, led by Koen Molenaar, offered an in-depth view of how Van Ameyde is reshaping the future of insurance claims.

A Process-First Approach to AI

Koen Molenaar emphasized a critical point from the outset: AI cannot function effectively without a well-structured process. Van Ameyde’s claims automation journey is rooted in a strong operational backbone, supported by their custom-built claim handling system, ECHO. Developed internally since 2006, ECHO acts as the central workflow engine that guides handlers step-by-step through the claims journey—from first notice of loss to final payout.

The Automation Toolbox: A Timeline of Innovation

Van Ameyde’s journey toward automation has evolved through several key phases:

• Straight Through Processing (STP) – Enables fully automated claims handling using business rules, particularly effective for low-complexity cases like glass or storm damage.
• Robotic Process Automation (RPA) (since 2018) – Deployed to automate repetitive, rule-based tasks across legacy systems, now supporting over 75 business processes.
• Process AI (since 2021) – Uses machine learning to interpret and categorize high volumes of emails and documents, reducing administrative workload by around 500 hours per week.
• Generative AI (since 2024) – Now being piloted to compose outbound communication, generate case summaries, and assist with intake and triage processes.

Real Impact: From Manual Intake to AI-Powered Decision Support

The most recent developments focus on an AI-native intake and triage domain, allowing automated processing of incoming claims. This includes:

• OCR-based data extraction from handwritten forms
• AI verification of document completeness and accuracy
• Automated email generation to request missing information
• Liability advice generation and initial reserve estimation
• Sketch and summary creation for claim handlers

This “human-in-the-loop” model ensures quality control while maximizing the efficiency gains of automation.

Proof of Concept: High Accuracy, High Confidence

Koen shared that Van Ameyde’s proof of concept for GenAI intake workflows has already achieved over 95% accuracy in specific domains, surpassing their internal benchmark for production deployment. This success paves the way for further rollouts throughout 2025.

Looking Ahead

The pace of innovation is rapid. With newer AI models capable of solving up to 75% of complex tasks—a significant leap from 35% just six months ago—Van Ameyde is exploring advanced orchestration of multiple AI agents across the claims journey.

As Koen noted, the future of intelligent claims handling isn’t just about implementing AI tools, but about reimagining the entire process to harness their full potential.

Watch the full webinar recording or connect with us via LinkedIn to hear about the next webinar.