social share alt icon

A TOP BANK IMPLEMENTED MACHINE LEARNING TO ENHANCE REVENUES AND CUSTOMER SATISFACTION

CLIENT

 

The client is an American, multinational, investment bank and financial services company. It needed a comprehensive payment investigation solution that could process unstructured data.

BUSINESS CHALLENGE

 

SWIFT messages in payment investigation cases frequently contained free text that was unstructured. While the data processing system of the client was able to analyze structured data, it was unable to process unstructured data.

SOLUTION

 

We, at Mphasis, introduced machine learning in client’s payment investigations to evaluate unstructured data for special instructions to automate processing and reduce errors. This approach also resulted in cost savings for the business. Our solution consisted of:

• Tools and technologies such as Amazon Web Services, TensorFlow, Python, and Fuzzy Sets which ensured that processing systems handled both the structured and unstructured data through natural language processing (NLP)

• A wide array of services from strategy, project management, design, data management, implementation, and testing

BENEFITS

Through our data strategy solutions, the client was not only able to process special instructions and improve financial savings, but it also gained other benefits such as

• Better operating margins by reducing the need for manual processes and encouraging human resources to spend more time on complicated investigations

• Improved customer satisfaction through shorter turnaround times for investigations resulting in happier customers and banking partners

• Increased revenues

• Enhanced quality and effectiveness as auditors spent more time on complex issues identified by machine learning, thus also resulting in greater effectiveness

• Reduced risk of inaction during investigations