Case Study: Topdanmark achieves 1–2 second homeowner coverage decisions with Domino Data Lab

Giving Homeowners Answers About Insurance Coverage in Seconds with Model-driven Policy Approvals

Topdanmark, a leading Danish insurance and pension company, faced slow, inconsistent manual reviews for homeowner coverage questions—customers who submitted photos and descriptions often waited hours or days. The data science team needed scalable, explainable models that could analyze images, text, and policy rules and be deployed and monitored reliably in production.

Using Domino’s platform, Topdanmark built model-driven workflows that combine image processing, natural language processing, and policy data to automatically evaluate coverage questions with reproducible experiments, rapid deployments, and real-time model monitoring. The solution delivers answers in 1–2 seconds (about 800× faster than manual review), automates 65% of cases, shortens approval times dramatically, and improves handling of model drift and complex cases.

Topdanmark

Stig Pedersen

Head of Machine Learning