Antworks Deriving Meaningful Insights From Unstructured Data

- Suchita Gonsalves

In today’s age of data explosion, businesses encounter the challenge of extracting value from the burgeoning unstructured data in the enterprise environment. The current goal of every enterprise is to segregate data with similar traits and classify them into clusters to make the best use of it. With the digital wave taking over, Machine Learning is proving to be particularly strong in solving the classification and clustering of data challenges. Machine Learning is being leveraged to make pattern-based decisions to automate processes and augment the accuracy of data-driven tasks.

One of the largest Title Insurers in the world set out to look for an intelligent automation provider to achieve its very data intensive goals. This company facilitates and streamlines real estate transactions by providing comprehensive title insurance protection and professional settlement services. One of their processes involved generating a title report after examining close to 49 different types of documents related to a property, in order to evaluate the chain of title sanctity. The daunting challenge was that the documents to be examined were highly unstructured documents - rendered not automatable with traditional OCR. To compound issues, the documents contained poor quality images due to dated records. Experienced examiners had to scrutinize documents using a highly recursive examination process resulting in very high handling times. These examiners had to go through intensive training before they could generate a title report.