Leveraging NLP to Dramatically Improve Customer Centricity
Companies have an opportunity to use NLP to dramatically improve their Customer Centricity. NLP can be used to systematically analyze vast quantities of language driven unstructured data sets like calls, complaints, or customer representative notes. Topic Modeling and Sentiment Analytics techniques have proven to be highly effective to identify, classify, and quantify customer needs, product innovations, customer experience enhancements, and customer servicing optimizations. More specifically we will discuss the effectiveness of fairly mature Topic Modeling techniques such as LDiA and some of the variations we have tested to optimize accuracy and applicability of results. We will also discuss the challenges and learnings of balancing unsupervised learning with human developed taxonomies and tagging.
Joan Gelpi has over 15 years of experience building and leading data science organizations in Fortune 100 companies. He has spent most of his career in Financial Services and other highly regulated industries such as Healthcare. He currently leads the Data Science organization at AIG. Prior to AIG he built and led a Data Science organization in American Express. He holds a PhD in Operations Research from Univeristat Politecnica Catalunya and an MBA from University of Chicago.