TUgis Conference 2019

Keyboards are on the verge of becoming “quaint” as more end users turn to conversational interfaces. These interfaces, or bots, offer the user a more convenient means of accessing and understanding data. For example, a user who is not adept at SQL would find it much easier to ask questions in a conversational style than to [perhaps] arrive at the same answer through syntax research and lots of trial and error. A bot eliminates this knowledge barrier and allows the user to ask more sophisticated questions of the data. It opens their imagination to all the questions the data might be able to answer and brings them closer to understanding the “Why?” instead of just “Who?”, “What?”, “When?” and “Where?”.

There are three major Natural Language Processing (NLP) platforms: Microsoft LUIS, IBM Watson Conversation, and Amazon Lex. In this talk we’ll focus on Amazon Lex and the ways in which a developer might want to incorporate it into their application. We’ll show how we added Amazon Lex to a mapping application and discuss the questions that informed our development decisions such as: “What type of end user might benefit from this service?”, “What should a user be able to do via voice command?”, and probably the most important question, “Is this a good use case?”.