Amazon Web Services (AWS) announced at its cloud summit in San Francisco that it’s rolling out Amazon Machine Learning, a fully managed, cloud-based service designed to pull useful information from mountains of data.
The problem with big data is that it often simply sits there unused because it’s far too complicated and energy- and time-intensive to find the critical information hidden inside.
AWS, following in the footsteps of cloud competitor Microsoft wants its new cloud service to help with that. Microsoft added a machine learning service to Azure in February.
“Amazon has a long legacy in machine learning,” said Jeff Bilger, a senior manager with Amazon Machine Learning. “It powers the product recommendations customers receive on Amazon.com. It is what makes Amazon Echo able to respond to your voice, and it is what allows us to unload an entire truck full of products and make them available for purchase in as little as 30 minutes.”
Machine learning, which is related to artificial intelligence, involves building algorithms that can learn from data.
Generally, machine learning is thought of as something used in robotics, to teach robot to navigate around a building or use tools. But companies like Ford and medical research institutes are increasingly using it to cull through big data to find patterns and connections not easily — or even possible — to suss out by humans.
Just last month, for instance, researchers at Carnegie Mellon University and the University of Pittsburgh announced that they are using machine learning to dig through prescription records, genome profiles, insurance records, diagnostic imaging and health records to help create treatment plans for people who not only have the same type of disease but share other similarities, like family history, active lifestyles and age groups.
One kind of cancer drug might work better on one person than another. The combination of big data and the artificial intelligence that can cull through it, allows scientists to develop designer treatments.
Now AWS’s Bilger wants to bring that kind of big data analysis to companies that might need to figure out what color sneakers sell better in New England, what kind of business process is the most efficient or what kind of social outreach creates the most loyal customers.
“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,” said Bilger. “Early on, we recognized that the potential of machine learning could only be realized if we made it accessible to every developer across Amazon.”
The idea is that with AWS’s new service, developers can use machine learning with the applications they build and run on the company’s cloud.
In an effort to make it easy for users to work with the data they already have stored in the AWS cloud, the new service is integrated with Amazon Simple Storage Service (Amazon S3), Amazon Redshift and Amazon Relational Database Service (Amazon RDS).
“It’s a cool thing and Amazon does know what it’s doing when it comes to analytics,” said Dan Olds, an analyst with The Gabriel Consulting Group. “Amazon counts on analytics to make its business model work. There are analytics working behind the scenes to predict what people might want to purchase next or to inform users what others have purchased. Plus, there are all of the back office analytics that tell Amazon decision makers how to best set up and stock the Amazon store.”
That kind of capability would help a lot of enterprises actually use their data. “The combination of machine learning and big data can result in companies gaining insights that they’d probably never have considered before,” added Olds.
Patrick Moorhead, an analyst with Moor Insights & Strategy, noted that while large enterprises could build their own machine learning system, using a cloud-based service would save them the massive expense, time and effort needed to build their own AI tools.
“When you combine the cloud, big data, and machine learning together, you get scalable capabilities to analyze and respond to a myriad of things,” he said. “With a service, you don’t need to procure, setup, find space for the hardware nor do you have to be an expert in datacenter software. You need to know the correct algorithms for measurement or find a way to get the data to AWS.