Kuwait Finance House
RiskGPT and Copilot for credit risk
Kuwait Finance House cuts credit-case risk rating time from about three days to under an hour
Problem
Credit case evaluation took three days on average and sometimes four to five. Reconciling aggregated data from subsidiaries took about two weeks.
Approach
The bank consolidated subsidiary data on Microsoft Fabric and built RiskGPT, an in-house risk engine, with an R&D institute and Microsoft AI services. Microsoft 365 Copilot and Power BI Copilot query the engine, and automated risk models run across branches.
Results
Takeaway
The data had to be gathered in one place before the risk engine could be used.
Categories
- Industry
- Finance and insurance
- Business function
- Finance, risk and legalProduction and quality
- Innovation type
- Process innovation
- AI technology
- Generative AIPredictive analytics and ML
- Adoption
- Co-developed with a partner
- Evidence
- Vendor's case study
Sources
Source summaries
The bank consolidated data on Fabric and built its RiskGPT engine. Microsoft reports that credit evaluation now takes under an hour, down from days.
Full details are in the original sources.