No, Microsoft’s new boss, Satya Nadella, hasn’t gone loopy on you.
But today the world’s largest software provider unveils a cloud-based, big data/analytics solution to help enterprises predict the future and to react to it — ahead of schedule — thereby increasing the odds of achieving a desired result.
“It’s been cooking for a while,” said Eron Kelly, general manager of the company’s data platform division.
“It,” to be more precise, is a Microsoft Azure Machine Learning service, AzureML, that helps enterprises apply advanced analytics to big data to forecast and take action on what is yet to come.
Filling the Data Scientist Gap and More
While you may be thinking that this isn’t earth-shattering news, that this is what data scientists have been doing for the last several years, the reality is that most companies don’t employ many (or even any) data scientists because there are so few of them to be had.
It’s not just that, but working with huge volumes of data takes time as does authoring, developing and coding machine learning solutions.
And then there’s the painstaking process of operationalizing trained models, load balancing, scaling up and scaling down ….
Introducing Microsoft AzureML
The magic of the cloud, Azure in this case, changes all of that, especially when it is leveraged in conjunction with the new analytics tools and some of the powerful algorithms that Microsoft has developed for its own use on products like Xbox and Bing.
“Our goal, when we built AzureML, was to make it easier for more people to participate in predictive analytics,” said Kelly.
So it goes to follow that with AzureML customers will be able to forgo the cumbersome tasks typically associated with starting Machine Learning projects and dive right into working with their data in easy-to-use GUI environments that include visual workflows, prewritten APIs, Web services and more. Data workers who relish working with Open Source R will find that it supports more than 300 of its packages.
Kelly said that AzureML was built for collaboration as well: “We want to make it easy for two or more people to participate."
Democratizing Big Data and Predictive Analytics
When AzureML is released next month, business analysts may find themselves able to generate the same kind of insights, with the same level of confidence, as data scientists.
Not only that, but Microsoft partners will also have access to its SDK (Software Development Kit) in order to build things like industry specific churn models which will be able to be implemented in short order.
From Data to Actionable Insights to Less Waste and Bigger Profits
We’ve all heard that “knowing your customer” is key to both bigger sales and stocking the right inventory, but could something like weather influence the kind of furniture people buy and when?
“Anticipating what customers will buy next is key to profit,” said Kelly. “You don’t want expensive inventory sitting around.”
At the same time, you don’t want to be caught short.
Predictive analytics are meant to solve these kinds of problems, as well as others ranging from predicting virus outbreaks to reducing energy consumption and many, many more.
Microsoft Brings Data Driven Fortune Telling to the Masses
We already know that Microsoft aims to bring the power of big data to a billion users. It’s clear now that they plan to bring the power of predictive analytics to a large subset of those users.
And, if everything beneath the front end works as well as the UI appears, they may have a winner. It certainly looks better and is easier to use than most others we’ve seen. And while many of the vendors we’ve spoken to claim to have “powerful” solutions, they were so powerful that we couldn’t even begin to use them without an expert by our side.
That’s a problem that AzureML doesn’t have, and it’s tackling some substantial data-intensive problems.
- Blame the C-Suite for Your Failed SharePoint Project
- Microsoft Leaks Offer a Glimpse of SharePoint 2016
- 5 Predictions About Marketing Technology
- Discussion Point: Who Has the Best Digital Marketing Hub?
- Why You Should Be Worried (and Angry) About Lenovo
- The Future of SEO is Not SEO
- Everything You Really Need to Know About Docker