Amazon Announces New AI-Based Alexa Developer Tools
- By John K. Waters
Amazon today rolled out a slew of new developer tools and features for its Alexa virtual digital assistant, including the beta release of Alexa Conversations, a deep learning-based dialogue manager designed to generate synthetic training data automatically from sample dialogs.
Released as a developer preview last year at Amazon re:Mars conference in Las Vegas, the much anticipated Alexa Conversations combines an AI-driven dialog manager with an advanced dialog simulation engine that automatically generates synthetic training data. The developers provide the APIs, annotated sample dialogs that include the prompts they want Alexa to say to the end user, and the actions they expect the customer to take. Alexa Conversations uses this information to generate dialog flows and variations, learning the large number of paths that the dialogs could take.
Alexa Conversations is designed to help developers quickly create interactions that are more conversational, explained Drew Meyer, head of Amazon's product marketing group for AI, Robotics, and Space, in a blog post, and not forced into strict patterns or sequential workflows. "In the past, developers scripted every potential turn, built an interaction model, managed dialog rules, wrote back-end business logic, and analyzed logs to test and iterate," he wrote. The Atom Tickets movie tickets skill, for example, used 5,500 lines of code and nearly 800 training examples. Built with Alexa Conversations, the app shrank by almost 70% to 1,700 lines of code, Meyer said, and it needed only 13 customer dialog samples.
With dialog samples provided by the developer, Alexa Conversations predictively models the dialog path using a deep, recurrent neural network, Meyer explained. "At runtime, this neural network takes the entire session's dialog history into account and predicts the optimal next action or step in the dialog, improving accuracy and reducing your design and code efforts," Meyer wrote.
Amazon announced a total of 31 new features during its Alexa Live voice developer event, including an API that allows the use of web technologies to build gaming apps for some Alexa devices, as well as a new service in preview called Alexa for Apps that's designed to let Alexa apps trigger actions, such as searches, within smartphone apps.
John K. Waters is the editor in chief of a number of Converge360.com sites, with a focus on high-end development, AI and future tech. He's been writing about cutting-edge technologies and culture of Silicon Valley for more than two decades, and he's written more than a dozen books. He also co-scripted the documentary film Silicon Valley: A 100 Year Renaissance, which aired on PBS. He can be reached at firstname.lastname@example.org.