Automating anomaly detection in legal documents for a leading Telecom Company.
Client Overview
Our client is a leading Telecommunication Services Company with operations in 26 countries and 142,000 employees worldwide. The Group has a global customer base of 259 million people, including 214 million mobile customers and 22 million fixed broadband subscribers.
The Problem
Client being a large telecom company had many legal documents that were greater than 1000 pages each. Client had a team that manually validated the legal documents and checked for errors and anomalies. But manually reading and validating the whole document is a tedious job for anyone and susceptible to human error. These highly sensitive legal documents were of crucial significance and the correctness of them could not be compromised.
The goal was to create an automated process that not only validates legal documents but can also identify errors throughout the document.
Our Solution
Our dedicated team at PureSoftware created a machine learning application to validate legal documents automatically using natural language processing. This application uses Semantic and format analysis to gain an understanding of the text so that it can achieve a higher level of accuracy in predicting anomalies. Applications’ machine learning abilities allows it to detect errors in documents, analyze them, and understand their patterns. This process helps to rule out similar errors and anomalies that may be present in the entire document. The application allows the system to get better at predicting errors with the passage of time.
The Result
The turnaround time of validating 1000+ page documents reduced from days to minutes.
100% detection of errors in the document.
The capability and accuracy to correctly predict anomalies improved over time.
