Last week the annual Gartner Magic Quadrant for Data Science and Machine-Learning Platforms 2018 was published. The old guard of SAS and IBM has tumbled this year with, Knime and RapidMiner taking the top spots in the Leader quadrant. Just as in years past, I expect the data science community to rant a bit about the rankings, who was included, excluded, added and dropped. However, this year the changes do seem to better reflect the reality of Python, R, Scala and Apache Spark in the data science world.

Honorable Mentions

This year I immediately noticed several significant data science players that did not participate or were omitted – DataRobot, Google Cloud Machine Learning and Amazon AWS Machine Learning. DataRobot publicly shared why they did not participate. They do encourage Gartner inquiries. DataRobot will also be showcased at Gartner’s upcoming Innovative Analytics in Action session. I don’t know why Google and Amazon are excluded. All three vendors did get an Honorable Mention in the report.

As the cloud landscape quickly evolves, we are starting to see new niche, open source and cloud players leapfrog traditionally dominant vendors. Ironically, Google was the only leader in The Forrester Wave™: Insight Platforms-As-A-Service, Q3 2017.  I have to imagine both Google and Amazon will be included in the future Gartner Magic Quadrant assessment. I’d be shocked if they did not want to participate. Then again, you never know.

Gartner could get disrupted in the future. Annual reports are practically outdated within a couple months in a cloud world with weekly releases. They did recently acquire crowd sourced, peer review companies. Gartner’s current business model depends on vendors paying them extremely high fees for advisory services, subscriptions, events, surveys, ranking report reprints, and so on. There has long been enduring controversy around ranking reports. They were publicly questioned again in 2014. While I do value the high quality of Gartner research and personally invest annually to attend Gartner’s Data & Analytics Conference, I also usually question the ranking reports. I don’t sense Gartner Magic Quadrants are rapidly facing extinction but it certainly could happen.

Interpreting Magic Quadrants

Before I get to the Data Science Magic Quadrant shifts, let’s revisit the topic of interpreting these reports. Keep in mind that it is always important to read the entire report to understand what is actually being scored. Do not confuse vendor placement with assumptions on solution capabilities.

Do not confuse vendor placement with assumptions on solution capabilities.

AtScale recently posted a fantastic video of Gideon Gartner, a founder of Gartner, that delves into known Magic Quadrant misuses and abuses. It is a fabulous find and a must watch! We will be analyzing the Gartner Magic Quadrant for Analytics and Business Intelligence Platforms 2018 report result together soon. In that session, we will further discuss this spicy, timely topic.

Gartner Magic Quadrant for Business Intelligence (BI) 2018: The Good, The Bad, The Ugly…


Top Movers and Shakers in 2018

Peeking at the 2017 and 2018 results, side by side and then overlaying them…, Knime and Rapid Miner clearly stand out this year. Alteryx and TIBCO also made great strides in the citizen data science space. IBM had the biggest drop – but – in the notes IBM’s Data Science Experience (DSX) offering was cited as having “potential to inspire a more comprehensive and innovative vision” leaving the door back to the Leaders quadrant within reach. SAS slipped but retained a coveted Leader placement. Other gainers include Anaconda, Domino, and Databricks. Other vendors that regressed include Microsoft, Dataiku, Angoss, Teradata and Mathworks.

Last year I walked through IBM, SAS, RapidMiner, Knime,, Microsoft and Dataiku one by one and shared my perspectives of those solutions. This year I’ll just quickly touch on a few different vendors since my free time is limited this week. Raising a rambunctious puppy while growing a small “gig economy” business has not been easy. I do love both but the cute furry one is winning more of my free time at the moment.

In my December 2015 Spark Big Data Analytics webinar, I introduced and demonstrated how to use the offering. I also highlighted this little company in my Industry Pulse magazine as a hot startup. Playing hands-on with, I found it to be easy to learn, quite efficient, and super powerful in the mega-hot Spark big data analytics ecosystem.

This year added Driverless AI automated feature engineering for machine learning. The machine learning automation concepts are similar to my personal favorite vendor in this space, DataRobot. Automated machine learning expedites development steps across the entire life-cycle. Projects that used to take weeks or months of effort can now be completed in hours. Do not fear automaton. These solutions empower data savvy talent to rapidly mine unprecedented levels of knowledge from oceans of data. Early adopters of automated machine learning tout unprecedented speed to insight and enhanced competitive advantage.


I was thrilled to see Anaconda here as I am a huge fan! The open source Anaconda Distribution is the easiest way to do Python data science and machine learning. I use Anaconda’s offering myself and also teach it from time to time in my Introduction to Data Science class. I love how Anaconda has simplified the download and management of numerous analytics tools, libraries and learning resources. It just works and it is fun.

Earlier this year I saw a white paper on Big Data Visualization with Datashader. The possibilities were truly impressive and something that I have had on my wish list to write about for a long, long time. Usually rendering big data visualization is limited at best. For my data visualization enthusiast audience, check out this white paper and Github samples.


Per Gartner’s notes, I do agree with their assessment. I reached out to this group several times with no luck. I’d love to do a solution review of Anaconda Enterprise and dig further into the premium offerings. I guess since they have grown to win over 6 million users, they don’t need to answer my email inquiries.


I was also pleasantly surprised to see the addition of Databricks Apache Spark-based Analytics Platform. In my summer 2016 Spark for Big Data Analytics series, I showcased several hands-on exercises with Databricks Community Edition. This is another vendor that seems to be doing exceptionally well and is not responsive.

The Databricks Unified Analytics Platform accelerates innovation by unifying data science, engineering, and business. Not only does it run an optimized version of Spark, offering 10-40x performance gains, but it also offers interactive analytics notebooks, integrated workflows, and enterprise security in a simple to spin up cloud solution. They also have a wonderful library of resources to learn Spark. That is how I found them.



TIBCO is an exciting vendor right now that is historically strong in middleware, event-processing, integration and business analytics. This one has been thrilling to watch grow deeper into Industry 4.0 connected intelligence with IoT and best-in-class streaming analytics. They are also expanding further into citizen data science with the recent acquisitions of Statistica and Alpine Data. In addition to making a debut on the Gartner Magic Quadrant for Data Science and Machine-Learning 2018 report, TIBCO also achieved strong rankings in The Forrester Wave™: Enterprise BI Platforms With Majority On-Premises Deployments and numerous other top industry analyst assessments.




Earlier this year I shared what was new in TIBCO Spotfire. I haven’t had a chance yet to share more from the TIBCO Systems of Insight portfolio. From a smart data catalog to inline data preparation, deeper analytics integrations with TERR/R,, ML/MLlib, SAS, MATLAB and the Statistica acquisition, StreamBase, live datamart, solution templates and accelerators, MDM, IoT, BPM, and on and on, TIBCO’s depth and breadth across the entire analytics spectrum has grown tremendously. They will need to integrate newly acquired solutions to better unify TIBCO’s portfolio offerings. If they do that, build on augmented intelligence and continue to shake things up with more unexpected moves, they should do even better next year.

Up Next

As usual, I will share a much deeper “analyzing the analysts” perspective of the 2018 Gartner BI and Analytics Magic Quadrant soon. That report was released late this year. Participating vendors are still waiting for permission to release statements and reprints.