Amazon Web Services this week introduced a new set of technologies designed to pull the artificial intelligence expertise developed for its core retail business into the cloud, while expanding its hybrid cloud technologies into the field, where data services can use the IoT to access connected devices.
AWS CEO Andy Jassy unveiled the new lineup of services at the company’s annual re:Invent conference, which drew a massive 32,000 attendees this year, demonstrating the rapid growth of Amazon’s increasingly important cloud services business.
Lex and Polly
One of the new offerings is Amazon Lex, a natural language technology that uses the same speech recognition capabilities found in Amazon’s Alexa digital assistant, which powers the company’s best-selling line of Echo home speakers and other devices.
Developers can integrate Amazon Lex into the Amazon Management console and use the technology to automate flight booking, or manage financial or other services via speech and voice recognition technology.
Along with Amazon Lex, the company introduced Amazon Polly, which allows developers to add natural-sounding speech capabilities to news readers or e-learning platforms. The technology offers 47 different male and female voice options and support for 24 different languages, facilitating its deployment around the globe. The Washington Post is using it to help create audio versions of stories and animated videos.
Amazon Rekognition is a tool that helps developers build applications for facial and object recognition. For example, Redfin, a full-service brokerage, is using the technology to help customers sort through millions of listings. SmugMug uses it to help customers manage billions of photos in storage.
“Amazon has absolutely used these new AI products in delivering its own services,” said Paul Teich, principal analyst at Tirias Research.
“I’m guessing here, but I’ll go out on a limb and predict that Amazon will soon start translating customer product reviews between languages, and then they’ll offer Google-like translation services,” he told the E-Commerce Times.
The company also announced AWS Greengrass, software that lets customers run AWS Lambda compute, messaging, data caching and sync operations on connected devices. A new data transfer tool, AWS Snowball Edge, allows data transfer rates at twice the speed of AWS Snowball (up to 100 TB).
The rise of the Internet of Things means that companies increasingly are managing devices devices that are powered by microprocessors with limited memory and processing power in factories, oil wells, cars and many other places. AWS Greengrass allows companies to add AWS Lambda functions to nearly any connected device, as if they were running directly from the AWS Cloud, facilitating operations without added downtime.
AWS Greengrass runs on nearly any device with a general purpose CPU using Ubuntu or Amazon Linux, according to the company. It supports ARM or x86 architectures.
Semiconductor makers including Intel, Qualcomm and Annapurna Labs are integrating Greengrass into their platforms so devices will have the technology built into their systems, AWS said.
Leader of the Pack
NASA’s Jet Propulsion Laboratory presented a demo at re:Invent showing how AWS Greengrass is working with Rov-E, the agency’s Remotely Operated Vehicle for Education. Technicolor is using AWS Greengrass for remote monitoring of set-top boxes, Connected Life tablets and other devices.
“AWS has always been the innovation leader in the IaaS market,” noted Jeffrey Kaplan, managing director of ThinkStrategies.
It “has put considerable pressure on its competitors by also providing the best prices and ease of use for its service,” he told the E-Commerce Times.
AWS “puts Amazon somewhat ahead of the pack,” said Michael Jude, a program manager at Stratecast/Frost & Sullivan.
“Amazon is one of the largest cloud services providers,” he the E-Commerce Times, “and as a result, any features they make available in the AI space will likely get a lot of traction.”
AWS also announced the launch of Amazon Athena, a serverless query service that allows customers to use standard SQL to access data from the Amazon Simple Storage Service, or Amazon S3, and get results in seconds.
The service negates the need to set up infrastructure or manage clusters, and because customers only pay per query, the company said, it makes the solution one of the most cost-effective options available.