re:Invent 2016 – New Amazon AI Services to Help Developers Build Smarter Apps Amazon Web Services revealed three new offerings Wednesday that will enable developers to create apps that leverage artificial intelligence (AI) and were available for purchase on Wednesday. During Wednesday’s keynote at the reInvent conference in Las Vegas, CEO Andy Jassy gave an overview of the Amazon Rekognition and Amazon Polly products. These products are part of the Amazon AI product group, along with the previously released Amazon Machine Learning. Amazon Rekognition, a sophisticated image recognition service, is now generally available. It can identify faces, objects, and scenes using labels (e.g. “car,” woman,” steering wheel), and can also identify people. It can detect certain attributes such as gender expression and distinguishing characteristics. Jassy said that it can face-match, which is a useful option to security app developers. Amazon Polly is also available. It is basically a text-to speech “parroting” service. The service responds to users by recording a.MP3 of the text. Polly can interpret real-like dialogue accurately. It can distinguish between homographs, and understand abbreviations. It can also “speak”, in 47 voices and in 24 languages. Amazon Lex is currently in preview of the three new AI services. Jassy stated that Lex is the technology behind AWS’ Alexa, Echo and other services and described it as a way developers can build natural language and conversational capabilities in their apps. Amazon Lex can be used to create chatbots and other types web [and] mobile apps that support engaging, real-life interactions. Jeff Barr, AWS evangelist, explained how bots can help you provide information, power your app, streamline work activities, or provide a control mechanism to robots, drones and toys in a blog post on Wednesday. This post coincides with the re-Invent announcement. You start by creating a conversation in the Lex Console and providing Lex with some examples phrases that will be used to build a natural-language model. Next, publish your Amazon Lexbot and let it process voice or text conversations with your users. “AI is becoming a reality because of the combination of better algorithms, broad access to large amounts of data, and cost-effective computing power offered by the cloud. Raju Gulabani, AWS’ vice-president of Databases, Analytics and AI, stated in a prepared statement that AWS hosts some of the most creative and innovative AI applications currently in use. Amazon has thousands of machine learning and deep-learning experts who have been working on AI technologies for years. These technologies are used to predict what customers may like to read, drive efficiency in our fulfillment centers using robotics and computer vision technologies, as well as to provide customers with our AI-powered virtual assistant Alexa. We now make the technology behind these innovations available to developers in the form three fully managed Amazon AI Services that are simple to use, powerful, cost-effective, and affordable. More information at re:Invent 2016.