Machine learning, sentiment analysis and linguistics all work together to create an AI-powered personal benefits assistant, Sofia.
Born from a hackathon at Businessolver in 2017.
However, customers were ultimately just not connecting with her or the experience. Clients did not find her to be a value-add, and often opted out of using her altogether. After mapping out all her touch-points we were able to conduct user research and usability studies around specific use cases, evaluate voice chats, and even in some cases chase the full experience by surveying users who choose to escalate their calls.
In one of the initial studies, we conducted it was reported that users found Sofia and the overall experience to be: “dated, flat, clunky, too technical, and having no personality.”
The bottom line: the conversational exchange was just not cutting it, and chats were escalating to call-center without either recording the data around the “why” or offering the user all available options in the dataset in a way that was not only digestible for the user, but also in a way that we as a data team could better document and quantify those triggered interactions to get more understanding around how to better engage the user. We were just not creating the right paths to facilitate the user experience successfully.
User Interface and User Experience Design for desktop product, browser based product, mobile and web.