Smarter Triage, Faster Claims: Transforming Subsidence Management for Crawford & Company with real-time AI
Using AI-powered triage and vision AI to create a first-of-its-kind platform that directly supports Crawford & Company’s claims process in real time.
THE CHALLENGE
Behind every insurance claim for Crawford & Company, is a customer relying on them to restore not just a property, but a sense of stability and peace of mind.
Subsidence is one of the most complex and costly areas of claims management. Traditional processes rely heavily on specialist consultants visiting properties to assess cracks and damage, then gathering supporting information, that can sometimes create bottlenecks and delays of months.
Crawford & Company needed a smarter, faster, easier and more consistent way to diagnose these claims – without losing sight of the human impact at the heart of every decision.
The Opportunity: Starting the Claims Process in Real Time for Faster Resolutions
What if Crawford & Company could begin assessing subsidence probability at point of first notification of loss and reach a reliable outcome within minutes.
From the moment a homeowner reports a potential issue, there’s already a wealth of data available: property location, soil type, tree coverage and weather patterns. Many of the factors that could contribute towards subsidence can be modelled using this kind of data, alongside historical claims and damage characteristics.
THE SOLUTION
We worked closely with Crawford & Company to design and build a digital platform that combines data, logic, and vision AI into one seamless process, including homeowner input to transform subsidence claims into a fast, intelligent, scalable solution.
This solution leverages AI-driven image recognition, structured data capture, analysed through our configurable decision engine for detailed algorithmic determination and next steps, to start the claims process in real time. As soon as a claim is created, the platform automatically pulls external datasets (soil, trees, weather, geology, historic claims) which are then combined with simple, easy homeowner-submitted evidence.
Within seconds, the algorithmic determination engine generates a probability score and decision recommendation, giving triage teams immediate, evidence-backed insight without necessarily requiring an expert consultant site visit.
Three core deliverables formed the solution:
Subsidence Probability Engine
A rules-based engine that calculates the likelihood of subsidence, drawing on multiple data sources:
Soil type, tree coverage, weather patterns, and property metadata
Historic claims data at postcode and property level
Homeowner-submitted evidence and photos
Geological and structural datasets
The engine outputs a clear determination report, similar to a credit score, supported by evidence that triage teams can act on immediately.
Homeowner App
A smart guided web app that enables homeowners to capture and submit structured
evidence with AI-assisted image recognition. With features such as:
Step-by-step photo capture for cracks, trees, and drainage
Smart prompts for common subsidence indicators
Seamless syncing with the probability engine
This shifts much of the evidence collection away from expert consultants, enabling
homeowners to capture critical evidence themselves, speeding up the process, with low effort for the customer whilst also improving consistency.
Internal Tools
To tie everything together, we built a web-based internal tool for Crawford & Company’s triage team, that centralises all data and generates clear outputs including:
Aggregated view of all third-party and user-submitted data
Traffic-light style triage decisions with full explainability trail
Visualised decision output from the engine
Full audit trail of submitted evidence
Integration with Crawford & Company’s case management systems
Together, these tools enable faster, evidence-led decision making while also reducing
reliance on site visits – all on a dashboard that integrates multiple datasets in real time for intelligent decision support.
THE RESULTS
By combining advanced AI, rules-based logic, and homeowner-led data capture, New Icon has created an industry-first platform for Crawford & Company that not only improves claims outcomes but also demonstrates what truly innovative digital claims management can look like.
The new digital triage platform marks a major step forward in how subsidence claims are managed, enabling faster and smarter decisions – balancing human expertise with intelligent digital tools.
Key outcomes:
Faster triage decisions - accelerating the process from the initial step in the claims journey to investigation readiness. For example, the first claim processed using the platform took just 72 minutes from submission to the customer to the customer providing all necessary information. Risk scores and evidence reports are generated instantly at submission, meaning claims handlers can begin decision-making the same day rather than waiting until an expert is able to make a site visit.
Improved decision accuracy – earlier detection of true subsidence cases with faster ruling out of non- subsidence cases and fast-tracking of more complex cases.
Better homeowner experiences – guided homeowner app reduces uncertainty, improving convenience, reducing friction and speeds up time to resolution.
Reduced costs – the new triage process enables claims handlers to make confident, evidence-backed decisions without always needing a specialist to visit the property - this has already led to a significant reduction in expert site visits, delivering substantial savings in both time and at an average saving of 11.4kg of CO2e, a far more sustainable solution.
Scalable innovation – algorithmic-based engine and vision AI provide a foundation for future automation.
