Building a Smarter Insurance Platform: UX, Data and AI in Action
Legacy Systems Blocking Data Access, Intelligence & Scale - how we reimagined Atradius Insights to deliver valuable data to their biggest customers faster than ever.
Legacy Systems Blocking Data Access, Intelligence & Scale
Atradius’ Insights platform was intended to provide clients and brokers with powerful visibility into credit exposure, policy decisions, and portfolio trends. Over time, however, the platform had become a bottleneck.
Issues with the Technology Stack
The Oracle ADF frontend and Exadata backend were end-of-life, presenting a defined risk to Atradius, as well as usability issues for their customers. Large enterprise clients were unable to run critical reports. The system would timeout after more than 10 minutes, meaning they couldn’t perform portfolio-wide analysis, creating a gap in their data and reporting. Coupled with slow load times, an outdated UX and UI, and disconnected workflows meant that Atradius’ clients weren’t able to make data-informed decisions.
At a time when customers expected real-time answers and smarter tools, Insights wasn’t delivering. Whilst a big challenge to overcome, the opportunity was clear.
Human Experience
Designed a new UI to match the data: fast, accessible, and clean.
Applied inclusive design principles, adhering to WCAG accessibility standards.
Seamless integration with Atrium, enabling users to transition from insight to servicing with no cognitive or platform friction.
Applied Intelligence
Developed an AI-powered search proof of concept (POC) where users can search in natural language: “Show me claims in the 6 months after the Suez Canal incident.” AI parses the intent and applies contextual filters automatically.
Enabled infrastructure to support predictive analytics, including risk trend modelling and generating six-month portfolio forecasts based on four years of user and sector data.
Atradius now has a platform capable of powering GenAI services like intelligent alerts, conversational insight, and embedded portfolio recommendation.
Systems AI-Readiness
Replaced Oracle Exadata with Azure MS SQL Hyperscale for on-demand scale.
Modernised ETL processes using Databricks, optimising for model-ready formats, low-latency queries, and future vectorisation.
Delivered a frontend in React, decoupling UX from backend complexity and enabling modular evolution.
Historical datasets spanning 15–20 years now load in under 5 seconds, even for top-tier clients.
Tangible, Quantifiable Impact
We shifted the lens from “make it look better” to “make it work, intelligently and at scale.” The numbers tell the story.
Most importantly, enterprise clients who previously abandoned the platform will now be able to use it daily to inform risk decisions. User feedback indicates that the new Insights platform is enabling faster, more confident decision-making.
Positioning Atradius to Harness the Full Power of GenAI
With legacy barriers removed and data pipelines redesigned, Atradius now has:
Structured, intelligent data flows
Supporting semantic search and predictive modelling.
A vectorisation-ready data layer
To enable future deployment of GenAI assistants, recommendation engines, or summarisation bots.
Design system and UI architecture
A design system and UI architecture that’s modular, evolvable, and secure—ready to support multi-modal AI interfaces.
This wasn’t just about rebuilding the Insights platform, it was about increasing the value of the platform to Atradius and its customers, becoming a strategic enabler for customer intelligence.
Conclusion
A Platform Built for Long-Term Insight
Atradius now has a modern insights platform that delivers fast, reliable access to critical data across the organisation. By consolidating reporting, improving performance and creating an AI-ready data foundation, the platform enables better decision making today while supporting future innovation.