If Amazon Web Services Inc. really wants to make machine learning services as ubiquitous as vanilla ice cream, its approach so far is turning it into something quite different. It’s more like Vietnamese coffee with frosted almonds and peanut butter curry, a flavor that can actually be ordered at one ice cream shop in San Francisco.

AWS announced a number of enhancements to its SageMaker machine learning development and deployment platform during the AWS Summit in New York City last week. Streaming algorithms? Check. Batch transform to pull direct from Amazon’s simplified S3 storage service? Just added. Improvements for Amazon Translate and Transcribe? That too.

The SageMaker iterations came barely three months after AWS rolled out a local mode feature for model training on notebook computers and integrated open-source deep learning tools into the platform. And SageMaker isn’t even one year old yet.

“Our mission is, we want to be able to take machine learning and make it available to all developers,” said Matt Wood (pictured), general manager of deep learning and artificial intelligence at AWS. “We want to make machine learning boring; we want to make it vanilla. It’s just another tool in the toolset of any developer and any data scientist.”

Wood spoke with John Furrier and Jeff Frick, co-hosts of theCUBE, SiliconANGLE Media’s mobile livestreaming studio, during the AWS Summit in NYC. They discussed the company’s three-tiered approach to its machine learning strategy, AWS’ evolving relationship with the open-source community, use cases for its DeepLens camera and recent customer engagement with the SageMaker platform. (* Disclosure below.)

This week theCUBE features Matt Wood as our Guest of the Week.

Countering open-source critics

The SageMaker improvements arrive at a time when the AWS strategy for machine learning is becoming more well-defined. The company has crafted a three-tiered approach, focusing on researchers and data scientists with open-source programming libraries, developers who want to leverage cloud services for data analysis, and application programmers interested in getting processes up and running quickly.

As the machine learning picture becomes clearer, so does the important role that open-source tools will play in the platform. When AWS announced its SageMaker improvements in April, the company revealed that it would add open-source frameworks, such as TensorFlow and MXNet.

“We provide a wide range of frameworks, open-source programming libraries that developers and data scientists use to build neural networks and intelligent systems,” Wood said. “There’s big, healthy open-source projects growing up around all of these popular frameworks.”

AWS’ willingness to embrace open-source marks an important progression for the cloud provider, which has taken its share of heat for not contributing enough to the community. The company has responded to critics by hiring open-source guru Zaheda Bhorat at the end of 2016 and joining the Cloud Native Computing Foundation last August.

In recent days, AWS has also been rumored to be considering an entrance into the data center switch market, leveraging open-source tools for what would be generic products. AWS denied the rumor through a statement released by Cisco Systems Inc. on July 18.

Aside from convenience, developers generally respond positively to the use of open-source tools in building machine learning models because they offer speed at scale. “The ability to pump in lots of data is one of the keys to building successful machine learning applications,” explained Wood, who described a decrease in training time between 10 and 25 percent. “We brought that capability to everybody that is using TensorFlow. This is a completely different way of how to think about training over large amounts of data.”

Use cases for DeepLens

AWS is also catering to developers through DeepLens, an AI-powered camera that is designed to enhance machine learning functionality. Announced in November, the camera began shipping last month at a price point of $250.

AWS is already seeing a number of use cases for the machine learning tool, according to Wood. These have included sending a text message to users when someone the camera doesn’t recognize approaches a front door and placing a book in front of the lens to trigger a text-to-speech algorithm for reading to children.

“With DeepLens, you can get up and running in 10 minutes,” Wood said. “It has been incredibly gratifying and really humbling to see developers that have no machine learning experience take this out of the box and build some really wonderful projects.”

Creative projects being generated by developers through machine learning will likely become more generally available over time, but the AWS technology is already being used by a number of high-profile customers. One of the announcements from the AWS Summit highlighted a decision by Formula One Group to use SageMaker for race strategies, data tracking, and digital broadcasts.

Major League Baseball, which had previously integrated AWS machine learning tools into its Statcast technology for player tracking, announced that the company would become its official machine learning provider.

“Where a lot of the momentum is going right now is SageMaker,” Wood said.“All of these groups are using the data, which just streams out of these races or these games. They can be for video or the telemetry of the cars or the telemetry of the players, and they’re pumping that through SageMaker to drive more engaging experiences for their viewers.”

There are eight additional AWS global summits scheduled over the next three months, and the company has shown no reluctance to date to announce new enhancements for its expanding machine learning platform. “There’s this wealth of opportunity ahead of us,” Wood said. “The responsibility that I feel very strongly is to be able to continually improve on that stack, to continually bring new capabilities to more developers.”

Don’t be surprised if those new capabilities come in many flavors, even vanilla.

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the AWS Summit in NYC. (* Disclosure: AWS sponsored this segment of theCUBE. Neither AWS nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE