Table of Content

CLIENT OVERVIEW

Our client is one of the leading Fortune 100 banks with client base of over a million people across countries. As an international bank, our client is bound to comply with numerous countrywide regulations and adhere to international standards as part of banking requirements and also to maintain their credibility as a trusted bank.

KEY REQUIREMENTS

As the information in all annexure documents was scanned manually, there always is a probability for higher risks of non-adherence to different lists like OFAC, KYC etc. due to human errors. Implications of non-compliance compelled our client to seek for a faster, trustworthy and an automated solution. A solution that can overcome challenges in complying with OFAC regulations which requires a thorough scrutiny of all the supporting documents involved in a particular transaction for pointing out suspicious or bad entity names. Following were the key requirements:

  • An automated solution to scan all the transaction documents accurately in order to reduce manual efforts and errors
  • A framework to identify entities and attributes compared against designated lists
  • A scalable system to handle large and growing volumes of data

KEY CONTRIBUTIONS

  • Designed a system which provided flexibility of accepting the input file in any file format including plain text, OCR text or HTML file by integrating seamlessly with client’s 3rd party OCR tools
  • Customized NLPro’ logic around 3rd party libraries to upload, classify, tokenize, process and mark the input documents in series of operations
  • Developed classification, decision making and machine learning algorithms for address detections and categorizing that text/data into different entities and re-using them for future checks
  • Licensed library’ default ad custom models and NLPro’s own custom logic, classification algorithms, lookup data, predefined rules and feature based address detection were integrated and customized as per the requirements
  • Implemented rule based text extraction and seamlessly integrated NLPro with client’ core systems, back office systems, Sanctions and OFAC lists
  • Implemented UI based training framework for applying incremental feedback for named entities
  • Provided ability to locate and classify elements into pre-defined categories that helped in quick analysis of the transactions / documents thus improving efficiency
  • Predefined a text template to support context based training using Machine Learning algorithms for increasing the scanning accuracy

KEY BENEFITS

  • Implementation of a framework based solution led to more than 95% accuracy of compliance checks
  • Significant reduction in manual due diligence work on financial documents ultimately reduced operational costs by 30%
  • Automation of the workflow process improved the productivity in document scanning and processing

NLPRO PROCESS FLOW

TECHNOLOGY STACK

  • HTML
  • JavaScript
  • JQuery
  • AngularJS
  • REST Services
  • File Polling (Batch mode)
  • XML
  • NLPro Framework