Employing AI to extract actionable insights from qualitative user feedbacks/comments

Everyday, thousands of DocuSign users leave qualitative product reviews via in-product surveys. These reviews represent a trove of rich data on user pain points and potential product innovations, but this information has been largely untapped due to the lack of a process for sifting actionable, product-specific comments from the larger dataset. In our presentation, we will describe how we built and deployed a multi-layer ensemble of models to organize and analyze qualitative user comments, allowing internal end users to drill down on relevance, sentiment, and topic.