Summary
A major global retail bank wanted to automate its document processing to yield faster and more efficient results. Evalueserve implemented a customized use of our IPDA capabilities to extract documents at scale to help the client meet this goal.
The Challenge
According to an IBM report, over 80% of all business data are embedded in different sorts of unstructured formats like emails, images, pdfs, and more. The highly regulated financial services industry is an especially document-heavy one.
As a large retail bank, the client spends large amounts of time and resources to process paperwork. Mortgages, car loans, credit cards, etc., all require various types of statements. Processing digital loan applications involves reviewing proof of income, bank statements, tax forms, and government-issued IDs. The client had to ensure all data was analyzed in an accurate yet efficient manner.
From start to finish, it can take non-digitally enabled banks a month to onboard customers. The client knew automation could shorten the process to just 3 to 4 days.
Our Solution
Evalueserve worked with the client to implement intelligent data processing and automation to accelerate loan processing and customer onboarding.
The solution started with Evalueserve’s cloud-native information extraction engine. The engine uses domain-specific AI designed and trained specifically for retail bank forms and documents. For further fine-tuning and accuracy, we trained the algorithms on actual client documents.
This engine helped the client classify unstructured documents and extract the required information from tables and paragraphs for credit decisions. Our domain-specific AI engine uses natural language processing (NLP) and machine learning (ML), allowing it to handle far more variability across documents and templates than traditional optical character recognition (OCR) and robotic process automation (RPA) techniques.
Business Impact
Evalueserve’s engine performs at 79% accuracy, which will increase as time allows for more machine learning. In addition, using our information extraction engine, the client was able to process more loans and onboard more customers. This directly increases profitability.
