PAW 2019 Conference Series – Mark Do Couto – Altair

Welcome to the 2019 Predictive Analytics World (PAW) Conference Series. Recorded live in Las Vegas, this series is bringing interviews straight to you from exhibitors and speakers at this year’s event. In this interview, host Jamin Brazil interviews Mark Do Couto, VP of Worldwide Sales, Data Intelligence at Altair.

Find Mark Online:

Email: mdocouto@altair.com

LinkedIn

Altair


[00:02]

My guest today is Mark Do Couto, Altair.  Mark, what do you think about the show so far?  

[00:08]  

It’s a great show.  It seems like there’s a good turnout.  Lot of great, interesting customers here.  It’s always good to see a lot of traffic and fellow vendors that are out here supporting.

[00:20]  

I really like how they laid everything out.  This is my first time at this particular event, and I feel like they’ve really nailed the cross-pollination between the attendees and the exhibit hall.

[00:29]  

Yeah, it’s really cool that some of the speaking engagements and stages are kind of opened into the exhibit hall; so, it makes a more natural flow.  

[00:37]  

Yeah, totally.  So, Altair – got it, right – is a recent acquisition or acquirer.  Tell me a little bit about the core business. 

[00:46]

Yeah, so, Altair, as a company, is traditionally a manufacturing and engineering company.  Focus on design, lot of time spent in the automotive, aerospace. But they really want to look into data and really get into the data space.  So Altair acquired Datawatch in December of last year. And Datawatch has a full suite of products, which goes straight from data preparation to predictive analytics to our own visualization software.  And most of our customers are based in financial services, which is completely outside of what Altair focuses on, which is a reason they were really excited. They were excited for us to stay in the market that we’ve been playing in, but also see how we can bring this into kind of the traditional manufacturing and engineering space.         

[01:28]

Oh, that’s interesting.  So it sounds it’s 0like a value-add to their existing, their core business but then maybe moving into another market?  

[01:35]   

Exactly.  So they, basically, have now acquired a company that already has customer and share in another market, but we’re having really interesting conversations with their traditional customer base:  a lot of helping predict machine failure, doing predictive maintenance on the manufacturing floor. So some really interesting use cases are coming out of their traditional customer base.   

[01:57]

That is actually really interesting.  Yeah, the downtime on, you know, if a cog goes out on a machine is catastrophic from an ROI perspective, right?

[02:06]  

Absolutely.

[02:07]

So being able to predict when machines are going to need maintenance and that kind of thing is actually a big industry that a lot of people don’t know about.

[02:15] 

Even some of the major automotive manufacturers that Altair traditionally works with, they have like trucking fleets.  So, if a truck is out on the road and it goes down, that delay can cost a lot too; so, we’re looking at ways that we can help predict when maintenance and support’s going to be needed on those trucks and get it done before it goes out on a cross-country haul.  

[02:36]

So, tell me about who’s your ideal customer?  Not, not..I mean Altair, obviously, is kind of this big but more in line with your core. 

[02:51]

Yeah, absolutely.  

[02:47]    

Datawatch or before.

[02:50]

So, the core of the business is really around how you’re leveraging the data that you have, but how you’re doing it in a way that you don’t need to be traditional data scientist.  So, really, it’s what Gartner would call your “citizen data scientist.” So, it’s a visual space to leverage data, build predictive models, and then automatically export that into a BI tool or even automatically generate code from that so that you can plug it into a deployment engine.  So our customer base is really people that are looking to leverage data science without having to get a Ph.D. in it. So it’s more of the “citizen data scientist,” but, in the same breath, if people who do have experience win R and Python, you can leverage that within our tool as well. So you can do some of that coding stuff if there are people on-team that want to do it.  But it’s all about keeping everything transparent. We could automate a lot of things if we wanted to, but we like to show all the different steps because that allows for anyone in auditing and governance to be able to see everything that was done, full transparency and then take it from there.   

[03:55]

That’s actually really interesting about the need to not build automation, right?  The not building automation actually adds a layer of value because you have an audit function that can pay attention to those steps. 

[04:10]  

Yeah, our traditional core business was always around financial services; so, we work with all of the major banks, the user software to some extent.  And for them, that was key. That whole black box solution or magic button just didn’t work for them. They really need to see the steps that are taken right from where the data source was taken in from to where it was deployed and gone live.  So we’ve tried to keep everything transparent as possible so all those steps can be seen. There are parts that we optimize. So we do a lot in terms of optimization to make the performance a lot quicker, but our customers seem to really like that.  

[04:46]

Oh, that’s really interesting.  Since you’re in financial services, BII got to be a big part of the whole value prop.

[04:52]

Yep, absolutely.  It’s one of the reasons why we ensure that our software can be run on premise within their environment.  We do have a cloud-based solution that we can provide but, at the end of the day, we let our customers choose how they want they want to deploy the software, and we meet all the security protocols to ensure that their data is safe.  

[05:10]

How long have you been with the company?

[5:11] 

Been with the company just over seven years now.  

[05:14]

So quite a while.

[05:14]

Yeah, absolutely.

[05:15]

Yeah, no kidding.  So you’ve seen a lot of…  I mean this whole democratization of access and data science is, from my vantage point…  It used to be really hard to run statistical models. I did it back in the 90s when I first started my career.  And it was a pain in the ass. I mean there was a lot to it. 

[05:34]   

Well, I think the advent of data growth, like the exponential growth of data, is leading us down a path of…  There’s this concept of sampling data and taking a small portion of it and building models off that and then applying that on the larger portion.  The problem with that is, if you have so much data, so many customers, how do you know that you’re properly sampling. So it’s one of the reasons why we’ve developed a piece of our software that runs native in Spark.  So, if a customer has a Hadoop setup, you can run this within Spark, leveraging all of Spark’s technology and actually build predictive models against billions and billions of records and not have to wait one or two days to process all that.    

[06:14]

That is crazy, right?  The volume of data that we’re dealing with now versus even five years ago is just so…  And you kind of look forward… I was talking to another guest before and he was talking about how intelligence is starting to get built into the product spec at the very beginning of the inception of the new product, right?  So, opposed to being this like more, “Oh, Gosh, now we need to pay attention,” it’s like built into… So now you think about the explosion that that’s going to have because it’s going to open up a whole bunch of streams of data that we otherwise, haven’t been able to access.     

[06:49]      

And to that point, it’s how are we looking at that and figuring out the best way to leverage that?  If we continue to leverage data the way we’ve been doing it for the past 10, 15 years, we’re going to left behind.  If the advent of collecting data, streaming data, bringing that data together, we need to jump on that train and get on that path and make sure that we can now analyze our data in the same breath as where the data’s being collected.  And I think that’s what we’re trying to do with this Spark kind of innovation that we’ve put together.

[07:16]

Are you more multi-sourced data inputs, meaning not.. maybe it’s market; maybe it’s internal customer.  I guess in a broad wrapper it’s BI, but customer voice, that kind of stuff. Are you seeing that?  

[07:33]

Yeah, everyone wants to create the ideal data lake or kind of the one point for data, but I can’t think of anyone of our customers that has data just residing in one data source.  Multiple data sources is the common thread. The question becomes how do you manage those multiple data sources. One of the components of our software is a browser-based data preparation tool that allows you to access all those data sources and be able to create your own data sets within this browser and actually share that with your team.  It gives you that collaborative format where you can actually see a team member that’s put a data set together and be able to leverage that data set throughout, maybe, a predictive model, for example. 

[08:19]  

Got it.  So, being a veteran in the industry, looking forward what do you think is going to be different in the next three years?

[08:25]  

That’s a great question.  The advent of machine learning and AI, depending on who you talk to, those terminologies could be marketing spin or whatever the case may be.  I think the biggest change that we’ll see is, as data continues to exponentially grow, there’s going to be new technologies that are going to be put in place to help leverage that data as much as possible.  And it’s how we’re leveraging access to that data and getting those responses as quickly as possible. At the end of the day, everyone wants to do something real time but, if data is becoming exponentially bigger, that real-time response is going to get slower.  So, there’s going to be a technology, I believe, that’s going to come out to be able to address this, and it’ll be kind of the next wave of what Hadoop was a couple years ago.   

[09:10]  

That’s super interesting.  That’ll be fun to watch.

[09:12]

Yeah, this is a great space; it has been a great space for years.  

[09:16]  

So, if somebody wants to get in contact with you ‘cause they feel like they have a project or question about your business, how would they do that?

[09:21]  

Yeah, they can just go to www.altair.com.  I mean they could email me directly if they’d like.  My email is mdocouto@altair.com and I’m happy to answer any questions that come up or get any of my team members to support.

[09:37]  

Mark, thanks so much for being on the Happy Market Research Podcast.

[09:39]  

Not a problem.  Thank you very much.

[09:40]

Hey, everybody else, if you please take the time to screen capture this, share in on your social media.  If you found any value at all, I’d really appreciate you taking that effort. I hope you have a fantastic rest of your day.  And special thanks to Predictive Analytics World, Marketing Analytics World, all the worlds wrapped up into one. You guys are awesome.  Thanks for hosting.