Technology Reply developed a solution for monitoring and managing banking institutions credit.


The project stems from the need of one of the main banking outsourcers, to offer a credit monitoring and management platform to banking institutions, in order to intercept abnormal scenarios, at bank portfolio level, and to be able to quickly manage them through actions aimed at stemming non-performing loans.


Technology Reply took care of the technological implementation of the functional requirements on the customer's big data platform, and of the entire data pipeline definition and development, from the ingestion stage to the data visualization and analysis stage.

The main components which make up the Early Warning monitoring ecosystem are:

  • A catalog of elementary indicators with dynamic parameters, from which the banking institutions managers can create complex alarm combinations to be activated on the credit portfolio, each alarm combination can be customized by each institute;
  • A backend engine, created using a data processing framework on big data clusters, which calculates and intercepts all customers meeting alarm parameters set by the bank, with a frequency that can vary from daily to monthly. This module aims to intercept and log, in a database, all the counterparties of the bank portfolio that have caused one or more alarms;
  • A frontend simulation environment, through which it is possible to view and simulate in real time what the effects of applying an alarm with specific parameters on the bank portfolio could be, in order to evaluate and choose the most appropriate parameters the institute, before scheduling the daily or monthly calculation;
  • A frontend data visualization module, developed through a data analytics platform, useful for outlining the historical trend and evolution of the counterparties that triggered at least one alarm. The module allows you to access details of each bank portfolio customer, in order to determine what the criteria were and which alarms intercepted him. In this way, it is possible to highlight risky behaviors which could generate NPLs.
  • A module for performance evaluation, useful for evaluating the overall performance of the entire monitoring system or alarms subsets, using KPIs which determine the accuracy, the efficiency and the impact.


Non-performing loans (or NPLs) monitoring, is of paramount importance within institutions which grant credit to third parties. The European Central Bank (ECB) considers NPLs monitoring one of the priority elements to be considered within a credit institution, due to their consequences on institutes final budgets.

Non-performing loans are scenarios where debtors of a particular form of credit, fail to repay it in full by its deadline. NPLs are divided by risk into various sub-classes, the final stage provides the debtor's "default" state, involving consequent legal actions.

Non-performing loans analysis is very important because loans weigh on banks income and resources, and this may limit their ability to make new loans and block the provision of further credit.

Therefore, it is essential to quickly intercept and resolve this cases, in a precise way, in order to avoid subsequent credit recovery procedures, which could require the bank further time and resources engagement.

Using a credit monitoring system it is possible, through the combination and application of simple rules, to develop indicators with a more or less fine granularity, with dynamic parameters, which allow you to register incoming alarms when one of the customers in the portfolio meets the established constraints.

In this way it is easy to monitor every day which and how many customers caused the activation of one or more alarms and, in case they remain on for a prolonged period, it will be necessary to initiate immediate actions that provide direct contact with the customer, through appropriate communication systems. This procedure makes the user aware of the monitoring of his status, inviting him to take actions to prevent the NPLs cases.