AWS promises to lower skills barrier to machine learning for developers

Amazon Web Services (AWS) has debuted a managed service for developers seeking a simpler way to incorporate machine learning processes into their apps.

The Amazon Machine Learning service enables developers to use the data they may already have stored in the AWS cloud to build predictive models that can be used to improve the way organisations deal with customer support, for example.

The technology has already been rigorously tested in-house at Amazon, where it has been used by its development teams to generate more than 50 billion predictions a week, the company claims.

The technology is used to offer Amazon.com customers product recommendations based on their past purchases, and is used to underpin its Siri-like voice command service, Echo.

“Amazon Machine Learning APIs and wizards guide developers through the process of creating and tuning machine learning models that can be easily deployed and scale to support billions of predictions,” AWS said in a statement.

The company also claims the service will make machine learning more accessible to the developer community by lowering the skills barrier needed to include it in the applications they make, because they will not require extensive knowledge of data analytics or specialist algorithms to use it.

Another key benefit of the service is that users will not need to worry about hardware provisioning, scaling or infrastructure monitoring, said AWS.

“Amazon Machine Learning makes machine learning broadly accessible to all software developers by abstracting away this complexity and automating these steps,” the company added.

There is no set-up cost for using the service, but AWS’s standard pay-as-you-go charging model does apply. Developers are invited to use the AWS Management Console and the firm’s APIs to create as many predictive models as they need.

Jeff Bilger, senior manager of Amazon Machine Learning, said the fact that the service has been used extensively in-house should put developers at ease about using it.

The company cited an instance where using the technology had enabled an Amazon developer to solve a problem that had previously taken several people 45 days to do.

“Early on, we recognised that the potential of machine learning could be realised only if we made it accessible to every developer across Amazon,” said Bilger.

“Amazon Machine Learning is the result of everything we’ve learned in the process of enabling thousands of Amazon developers to quickly build models, experiment, and then scale to power planet-scale predictive applications.”

Outside of AWS, the service is being used by global media and entertainment company Comcast and mobile gaming startup Space Ape Games.

Toby Moore, CTO and co-founder of Space Ape Games, said his firm has used the technology to help shape the type of content it pushes out to users of its apps to ensure they keep playing its games.

“By using a service like Amazon Machine Learning, we are able to make decisions more easily and precisely about how to keep our customers excited about and enjoying playing games,” he said.

“We plan to use Amazon Machine Learning across multiple departments in our organisation to help us build and deploy predictive models for our current and future games.”