Automated financial spreading and credit evaluation for a USD 60 billion systematically important financial institution.


The client, a USD 60 billion systematically important financial institution with operations across multiple geographies, had an entirely manual process for financial statement spreading and credit evaluation, a significant task considering the size of the bank and the large scale of its operations.

Automating the financial spreading process and centralizing all credit evaluation processes onto a single platform was no longer a choice - it was a necessity given the fast-paced digital transformation taking place in peer-group banks and financial institutions.

Senior client stakeholders sought help in developing an intelligent, proprietary scoring mechanism to automate and standardize the analysis of financial statements, and in implementing a centralized platform for the bank's credit processes. These initiatives would allow the client to build comprehensive customer profiles and accelerate the process of providing credit limit approvals.


Under the existing system, a large team of research analysts was engaged in carrying out credit evaluation using manual, repetitive processes. Such a method of financial statement analysis and credit appraisal was

  • Error-prone and inefficient due to manual input of data from hard-copy documents
  • Inconsistent since it was difficult to maintain standardization when human effort was involved
  • Cost-ineffective due to fixed and recurring costs attributable to a large team setup

Manual processes led to delays in the entire credit appraisal lifecycle, from data collection and analysis to credit scoring and approvals. The result of this was a steady backlog of cases and high turnaround times.


Extensive discussions were held with the client to understand the existing process and identify areas for digitization, automation and standardization. The ScaleCred platform's proprietary pixel-based machine learning technology was ideally suited to undertake all these activities.

With the required level of technology customization, the following processes were undertaken:

  • Document digitization and uploading services were set up to convert hard-copy documents into electronic form.
  • A pixel-based character recognition and artificial intelligence (AI) platform was implemented to automatically extract information from scanned documents.
  • Deep learning algorithms were trained to analyze data from a variety of scanned documents, many of them with high 'noise' levels.

Digitization and automation initiatives resulted in the following:

  • It became possible to standardize information across the bank based on a proprietary scoring methodology, something that posed a significant challenge under the manual system.​
  • Scanned report contents could be viewed in text format, allowing financial data to be automatically exported into Excel templates and APIs.

Machine learning-based automation, a centralized platform and proprietary scoring methodology quipped the client with the ability to

  • Evaluate and compare financial information in a standardized format.
  • Get a complete view of customers' financial profiles on a single dashboard.
  • Calculate risk scores/credit lines on a real-time basis, enabling faster decision-making.

The platform also enabled the delivery of real-time alerts on latest industry developments.

The ScaleCred solution automated manual, repetitive tasks and helped in freeing analyst bandwidth to take up more analytical functions. Standardized data and a proprietary scoring methodology led to objective and faster decision-making.