Data environment.
“Tier one bank required a modelling data environment that made all available credit risk data; discoverable, accessible, standardised and usable for Data Scientists. Prior to engagement the approach implemented to gather requirements and develop the solution was not compatible with the complexity and uncertainty of the user requirements. As a result, the project had spent almost half its budget and delivered no real value. QRisk specialists were engaged to rescue the project and secure its original objectives.”
-
CLIENT PROBLEM
Tier one bank required a modelling data environment that made all available credit risk data; discoverable, accessible, standardised and usable for Data Scientists. Prior to engagement the approach implemented to gather requirements and develop the solution was not compatible with the complexity and uncertainty of the user requirements. As a result, the project had spent almost half its budget and delivered no real value. QRisk specialists were engaged to rescue the project and secure its original objectives
-
SOLUTION
Several approaches needed to be implemented as part of the project turnaround and included:
Requirements stocktake and clarification using working groups by subject area; customer, arrangement, default, collateral, and ratings.
Delivery process improvements using agile delivery as a mainstay for the project turnaround.
Technical capability improvements utilising automated testing capability.
-
CONCLUSION
The modelling data environment is now a well established and governed BAU platform for model development, validation and monitoring teams across the organisation.
Reduction in model lead time by 25% - 40% (3-6 months) depending on the model and activity (build, validation).
Enterprise data is now discoverable, accessible, standardised and usable for multiple teams to share.
More efficient processes and reducing the operational risk of data errors.
The project turnaround successfully delivered the required functionality within the initial project budget.