API World + AI DevWorld

There is a growing need to make big data more accessible and intuitive for enterprises. By applying artificial intelligence to power smart conversations about data, chatbots will transform the way companies rely on intelligent conversations to make decisions. Having a personal analytics assistant makes it as simple as having a conversation to get all the powerful answers you need for success. A main take away of this session would be the advancements in voice first technology and how natural language processing is shaping the way we understand and use data. Voice, after all, is the most primitive form of communication.This talk focuses on an innovative process to capture a database’s metadata and the end users’ vocabulary, and use definitive queries to rank distances using a word embedding approach, to deliver optimal responses to questions in natural language. To support this level of personalization, semi-automated disposable models can be implemented rapidly to support questions on proprietary databases from many different points of view - transactions, devices, customers, tickets etc. These models can be improved continuously through retraining based on user feedback. This flexibility supports an iterative process for developers to implement changes on a daily basis when necessary - instead of using one massive AI model that takes too much time to build and is worth too much to dispose of.