I’m here with George Mathew, President and Chief Operating Officer of Alteryx. The Inspire 2014 conference just wrapped up. George thanks for taking a few minutes to chat with us.
Absolutely. Thanks so much for having me. I appreciate it very much.
Can you give us an introduction to Alteryx so our listeners get a nice picture of what the product is?
Sure. Alteryx is really focused on this idea that data analysts should have better capability in their hands to blend data and run advanced analytical models in that same user experience that you would almost look at the visual analysis being done in Tableau.
Really that’s what Alteryx is focused on, this idea that you should be able to get a user experience in the hands of users that are not necessarily programmers day in and day out, to be able to very quickly blend data, to be able to run data quality functions, data integration profile and functions and then be able to run analytical models – predictive models, statistical models, algorithmic models, spatial models. We then make it very, very straightforward to then package that output directly into an analytic application or into a visual analysis experience like Tableau or Qlik.
I love that Alteryx hits on a couple of different really key points in the whole process of working with data. You just released version 9.0 a couple months ago and I’m really intrigued by some of the new capabilities. What do you think is the most important problem you solved with this release and how do you see it changing the way companies will work?
One of the things that we’ve been really pushing the envelope on is this idea that you have to have a useable software experience in terms of any data processing and data analytic work that folks are doing. And so, when you look at 9.0 itself I think there’s sort of two major themes that really come out as really pushing the envelope of how that user experience plays itself out.
One is this notion that there are more and more forms of information, particularly on customer analytics and customer insights, that should be in the hands of users. And so, when we looked at the 9.0 release we went ahead and supported social media data, for instance being able to take aggregations that come from social media providers like Datasift and being able to directly go against the social streams of information that Twitter or FourSquare provide. This is something that we felt is really important to expand the usage of Alteryx, particularly when it comes to customer analytics.
The second thing that we felt was really compelling in this experience is this idea that I mentioned earlier, which is how do you get those insights shared to a broader set of users? We really have this notion of an analytical app. How do you build and how do you basically push an analytical app into a broader set of consumers? Here, Alteryx did quite a bit of work in terms of making the user experience of building an analytic app just as seamless as you would work with the data and the flow around data in the Alteryx designer.
So I think these two real concepts of 9.0 in terms of one, expanding the experience for anyone who’s doing customer analytics. Then two, making it easier for them to create consumable analytical applications in one design time. It’s a great differentiator for Alteryx but also just the release of 9.0 into the market.
Those are very cool ideas. The notion of an analytical app is fairly different than the way most people talk about working in this industry. Now, you’ve got these two big ideas and started to solve them in version 9.0, what’s on your mind about the next big challenge that you’d like to tackle? But I have a little bit of twist to that. What do you think has to happen outside of Alteryx to make that idea possible?
Sure. I think one of the things that I’m seeing pretty closely right now is how much data is becoming accumulated particularly inside of Hadoop as an infrastructure and landscape for data management. This is actually happening almost at petabyte scale.
We at Alteryx believe there’s a real opportunity for our users to take advantage of these data lakes that are effectively formed using Hadoop as the underlying infrastructure. One of the next big things for Alteryx is really how much more we can very seamlessly work, with not only those data lakes, but be able to take advantage of the compute and storage that’s effectively available in Hadoop.
For years, we’ve always had support for direct integration via a query based interface where you can go ahead and have a SQL query go in via Hive or Impala and be able to structure data out of the Hadoop landscape and be able to pull it into your analysis. That works today. Our customers use it and they love it.
One of the things I mentioned as far as the follow on to the 9.0 release, is the ability to go directly against the file system itself, the HDFS file system. This enables you to read and write semi-structured and structured data directly into the Hadoop landscape. But over time we really see this idea that there’s more and more interactive, real-time engines that are available to compute against data that’s effectively stored inside the Hadoop landscape. This is where we see ourselves really, really close to a project known as Apache Spark, and really taking on the ability to make that compute very portable between whatever happens on the analysis on a design time like the Alteryx designer, all the way into scale out, clustered computing that occurs in the Hadoop landscape. We’re really excited by that direction, and where the Alteryx and the Hadoop communities are headed.
Look for Part 2 in a couple days …