Case studie: Invex

Supporting Invex on its path to becoming a paperless bank

Steps to Eliminate Paper in Bank Operations

Initial Challenges and customer Context

Towards a More Agile and Digital Banking

The change began with a question: What if we go paperless?

1

Identity cards, proof of address, income statements, account opening forms, contracts, affidavits, bylaws, powers of attorney, and financial statements: these are some of the documents that banks require from their customers as part of the identity verification and Know Your Customer (KYC) process, necessary for onboarding and approving credit applications or other products.

At INVEX, the data from these documents had to be manually entered into internal applications, which meant that administrative team members had to copy and paste information from paper documents or, in the best-case scenario, from previously digitized files. These highly manual and tedious tasks were particularly prone to errors.

The information extracted from these documents could be structured (such as an identity card), semi-structured (like an invoice), or completely unstructured (like a contract). This diversity posed a significant challenge since unstructured documents are not ideal for traditional automation based on RPA.

Aiming to become a paperless bank, INVEX requested a solution from EDSA based on Artificial Intelligence to automate the processing of these documents.

Paperless projects provide significant benefits for organizations, such as promoting sustainability, saving space and archiving costs, increasing transparency, reducing the risk of information loss, enabling agile and remote access to data, and most importantly, enhancing the customer experience by speeding up processes and allowing faster delivery of products and services.

Solution Development

Intelligent Automation to Unleash Operational Potential

We Applied Artificial Intelligence and RPA to Automate Document Processing.

2

Transforming INVEX's traditional processes and work methodology was a challenge for EDSA, especially when incorporating Artificial Intelligence into the existing registration processes.

Manual document handling involved laborious tasks such as opening individual images or PDF scans, analyzing files, and entering relevant data into applications. After a process of testing and evaluations, EDSA recommended implementing UiPath Document Understanding, an AI-based solution for reading, extracting, validating, and exporting data from digital documents.

The project began with a process discovery in the Trust Division, characterized by its high use of paper. All involved documents were reviewed, their taxonomy recorded, and relevant data identified and structured so that the tool's OCR (Optical Character Recognition) engine could accurately identify and extract them.

As is common in such projects, the initial reading was not always correct due to unstructured data, legibility issues in scans, among other factors. To address these limitations, the tool implements the concept of 'human in the loop,' allowing users to validate readings, correct deviations, and, thanks to the Machine Learning model, automatically train the tool to improve its accuracy.

In addition, EDSA developed enhancements on the UiPath Document Understanding framework to allow simultaneous processing of multiple documents (originally processed one at a time), increasing automation capacity and optimizing overall system performance.

Impact and Achieved Benefits

A Paperless Bank: Faster and More Customer-Centric

Less paper, fewer errors, more focus on what really matters.

3

In less than a year, INVEX managed to automatically process around 500,000 documents in areas such as Trust, Customer Service, Credits, among others.

Although the project began as an initiative of the Trust Division, the benefits quickly extended throughout the organization. Other departments identified that they could reuse many of the already processed documents, such as passports, identity cards, corporate forms, and certificates issued by government entities.

With this experience, INVEX encouraged its teams to be trained in the use of UiPath Document Understanding, allowing them to autonomously manage document taxonomy, correct deviations, and train Machine Learning models.

Thanks to this project, INVEX achieved:

  • A significant reduction in paper usage.
  • Accelerated information processing.
  • Reduced data entry errors.
  • Enhanced customer experience.

A transformation that not only optimized internal operations but also reaffirmed INVEX's commitment to innovation and service excellence.