Talk-To-Your-Data for Reinsurance

Supercharge user access to research data with StatGPT and Mind Maps.

Request Demo

4x

Faster access to cross-dataset information

90%

Quality rate of automated data search

2x

Increase in self-service search by business users

Challenge

Swiss Re has been providing digital research analytics and risk predictions to customers for over 25 years. EPAM Partners with Swiss Re to build a modern, AI-Powered Data Portal that enhances operational efficiency, data exploration capabilities and increases market impact.

Industry

Insurance

Dev team size

35 developers

Duration

10 months

Solution

Working with Swiss Re, we identified the main pain points - the client provided unstructured PDF articles and structured statistical data, both of which could be cumbersome for end-users to access. The goal was the revamp their portal and provide natural language access to both the publications and underlying data.

The DIAL team developed a data ingestion pipeline and imported all of the existing publications and time-series data into QuantHub, a full-scale solution for creating, processing, and publishing data products. This enabled leveraging StatGPT, an agentic application that allows users to talk to the research publications and retrieve the time-series data from one simple chat-based UI.

Using DIAL RAG, the publications search from the portal was improved substantially, allowing users to find the research they are looking for more effectively.

The content of publications was also turned into an interactive knowledge graph powered by DIAL Mind Map Studio, enabling users to visually navigate through the publications, and also ask questions about them in a chat-bot.