UTP Collection Engine

The Collection Engine project was born as part of an agreement between a main Italian bank and a servicer, a company providing credit recovery services, for the assignment to the latter of a UTP customer portfolio on which to carry out debt recovery credit. The purpose of the application is therefore to calculate collections, recovery performance and fees to be paid to the servicer for the management and recovery carried out.

The engine calculation perimeter was defined as follows:

- UTP clientele

- Corporate Customers

- Customers both in internal bank management and in servicer management

Functional components

The application carries out massive processing in order to calculate collections, recovery performance and fees for the servicer:

• Data Ingestion: recovery of sources from several 10 different sources within the Bank

• Data processing: Calculation of collections, actual and target performances, and fees

• Data output: Create streams to feed other legacy applications with the calculated indices.

Technical components

The application is divided into 5 technical components:

• Spring Batch-based Batch Engine for Data Ingestion, data processing and file creation operations to feed downstream systems

• Spring Cloud Data Flow orchestrator to allow the implementation and scheduling of job chains

• BackEnd component in Java 11: exposes data saved to the DB to the interface, creates reports, manages adjustments on data created by users

• FrontEnd component in Angular 14: allows the user to query data and enter adjustments

• Oracle database: cornerstone and data consistency point of the entire solution; it stores the data of the ingestion phases and offers a scalable layer on which the processing of the indices necessary for the bank is structured

The technological core of the solution is based on Spring Batch and Spring Cloud Data Flow. We chose this framework.

We chose a Spring Batch and Spring Cloud Data Flow solution because:

- Allows integration with Openshift RedHat on an on-premise cloud

- Handles large files (up to 100 GB)

- Offers the possibility of multithreading, allowing you to manage variable workloads effectively

- Allows you to schedule complex chains, in a flexible, dynamic way facilitated by a graphical management dashboard

- Supports horizontal scaling

The Collection Engine allows the Bank to improve the following aspects:

• Operational Efficiency: The application allows you to automate complex processes and calculations, reducing the time and resources needed for case analysis.

• Data Accuracy: The solution minimizes human errors in recovery calculations and related performance

• Reporting and Monitoring: The application provides detailed reports on the recovery progress, facilitating the management and optimization of strategies.