BUSINESS CONTEXT

Financial Decision Engine is a specific service integrated into the existing IT infrastructure of the bank. It makes decisions based on the data collected from different areas. The service functionality includes:

• Checking, approving and rejecting loan applications
• Conducting the calculation of credit parameters for each specific application
• Implementing additional checks for executable contracts at different stages of their lifecycle
• Determining the availability or unavailability of other banking products for specific customers

CHALLENGE

Aterise team was required to build a service that collects a considerable amount of data from diverse sources. Some of them are internal bank products, and some are external partner services, providing API

The set of decisions that the service should take differs a lot. Moreover, various solutions should additionally be configured by a set of parameters, which depends on many factors

The work of the service should be completely transparent for bank employees who should see the details of the decision made, the course of the decision itself, the reason for refusal or approval, and, if necessary, provide details to the regulator on all decisions taken in the past

The customer wanted the service to seamlessly fit into their existing infrastructure

SOLUTION

We developed a micro-service which satisfied all the customer’s requirements. The full life cycle of this service was successfully integrated into the internal product delivery processes as well as into its infrastructure. Our developers configured within change requests several types of solutions that are integrated into the client’s roadmap release cycle. This allows the system to be stable and meet the regulator’s requirements

The inside-the-service flexible mechanism for creating solution templates now allows implementing flexible parameterization depending on the environment, service configuration as well as on features of the input data for this particular solution. This mechanism allows us to make changes to the already configured types of solutions as well as create new solutions with the lowest possible cost

We also implemented a storage mechanism for decision-making that most accurately reflects all the peculiarities, decision progress, and all data that was obtained or calculated, as well as the user’s interface module for the back office of the bank that displays this data in a human-friendly form. This mechanism allows us to see new types of data and solutions while keeping historical data unchanged

The service was successfully released and is demonstrating the stability and reliability of its performance. The implementation of this service has resolved one of the most critical points in building a grid of fully online products. During the first 6 months, the system has processed more than 1000 solutions per day and still continues to grow