Summary: Forrester has just released its “New Wave™: Automation-Focused Machine Learning Solutions, Q2 2019” report on leading stand-alone automated machine learning platforms. This is our first good side-by-side comparison. You might also want to consider some who were not included.
You know you’ve come of age when the major review publications like Gartner and Forrester publish a study on your segment. That’s what’s finally happened. Just released is “The Forrester New Wave™: Automation-Focused Machine Learning Solutions, Q2 2019”.
This is the first reasonably deep review of platforms and covers nine of what Forrester describes as ‘the most significant providers in the segment’. Those being Aible, Bell Integrator, Big Squid, DataRobot, DMway Analytics, dotData, EdgeVerve, H2O.ai, and Squark.
I’ve been following these automated machine learning (AML) platforms since they emerged. I wrote first about them in the spring of 2016 under the somewhat scary title “Data Scientists Automated and Unemployed by 2025!”.
Well we’ve still got six years to run and it hasn’t happened yet. On the other hand no-code data science is on the rise and AML platforms along with their partially automated platform brethren are what’s behind it.
Here’s What Forrester Found
Let’s cut to the chase. What everybody wants to know is who are the winners and losers in this race. Here’s Forrester’s graphic.
A big shout out to Forrester for being first to recognize the significance of this segment. However, it’s important to clearly understand how they picked these participants. Candidly, I’d never heard of some of these although I’ve been trying to keep up.
First of all a Forrester “New Wave” report is different from their usual in-depth studies. Participants were invited from among those Forrester thought were leaders and had to submit an RFP like response. Research was limited to a 1 ½ hr. interview / demo and calls to some of their installed customers.
Inclusion criteria required the vendor to have at least five installed customers and have a full spectrum AML offering. That starts with data blending, prep, cleansing, feature extraction, feature engineering, model creation, and includes model deployment. That’s a full soup-to-nuts offering.
So the vendors who participated were to an extent self-selected and there’s no indication of who might have applied and been rejected. I suspect there are a number of other competitors who could have passed these hurdles.
The grading criteria seem quite good. With 10 categories, 7 are aimed at functionality and 3 at strategy. Here’s how the grading ended up:
It came as no surprise that DataRobot was in the lead. They have invested heavily in both functionality and in market share. I’m guessing they have the largest installed base.
H2O.ai is also a well-established company but one that made a remarkable pivot to AML only in the last two years. Previously they were a very well regarded DL tool for data scientists and very code oriented.
Of the group, as you might expect, Forrester points out that some are making their case mostly to data scientists and others mostly to non-data scientist. Keep in mind that this last group, the so-called citizen data scientists includes both LOB managers and more significantly the large cadre of BI analysts. It’s an interesting study and worth your read. Google it and you’ll be able to find a copy as I did.
Those Left Out
I think this is a great study but I’m not sure I would make purchase decisions based solely on its contents. Forrester acknowledges that they did not include the cloud based AML offerings from the likes of Amazon, Microsoft, Google and others. They promise to review them in an upcoming study and given the power behind those companies, we expect to see really excellent functionality.
Second, there’s a long list of stand-alones that might be added to this list. There’s no way to know if they were invited and didn’t make the cut, or simply missed in the invite. Based on my reading over the last few years my guess is that there are at least 30 AML platforms in some early stage of release.
This is not an endorsement, but I was surprised for example not to see some of the following included:
- Xpanse Analytics
Some of these have been around for a long time, and some even offer automated deep learning as part of their platform if you’re ready to move in that direction.
Then there’s the whole category of open source. Like all open source this will appeal to those of you willing to roll up your sleeves and commit to some DIY. See the list in some of our previous articles (links below).
Finally there are the augmented or assisted platforms that are greatly simplified but not fully automated. Stalwarts like SAS, RapidMiner, IBM SPSS (ne Watson), Alteryx, and BigML are in this category.
So when I qualified the coming of age story with the ‘almost’ my meaning was that there are some platforms that are enterprise-ready today but there are still a bunch of fast followers who may also be of interest to you.
I look forward to a full-on in-depth review by Forrester or Gartner of this segment. AML is coming. You could jump in now with one of these leaders but there will likely be interesting new developments over the next year or two.
Additional articles on Automated Machine Learning, Automated Deep Learning, and Other No-Code Solutions
Practicing ‘No Code’ Data Science (October 2018)
What’s New in Data Prep (September 2018)
Democratizing Deep Learning – The Stanford Dawn Project (September 2018)
Transfer Learning –Deep Learning for Everyone (April 2018)
Automated Deep Learning – So Simple Anyone Can Do It (April 2018)
Next Generation Automated Machine Learning (AML) (April 2018)
More on Fully Automated Machine Learning (August 2017)
Automated Machine Learning for Professionals (July 2017)
Data Scientists Automated and Unemployed by 2025 – Update! (July 2017)
Data Scientists Automated and Unemployed by 2025! (April 2016)
Other articles by Bill Vorhies
About the author: Bill is Contributing Editor for Data Science Central. Bill is also President & Chief Data Scientist at Data-Magnum and has practiced as a data scientist since 2001. His articles have been read more than 1.5 million times.
He can be reached at:
[email protected] or [email protected]