Data Masters Podcast
July 20, 2023

Harnessing the Power of Data Analytics for Business Innovation: An Interview with Adam Wilson, SVP and GM at Alteryx Analytics Cloud

Adam Wilson
Senior Vice President and General Manager at Alteryx Analytics Cloud

In today's highly competitive business environment, the key lies in discovering new and interesting data, seamlessly onboarding it, and stitching it together to form training sets that generate powerful algorithms. Accelerating this process goes beyond operational savings. It's a game-changer, allowing you to outpace your competition, deliver unparalleled customer experiences and capture greater market share.Join us today as we welcome Adam Wilson, Senior Vice President and General Manager at Alteryx Analytics Cloud and former CEO of Trifacta before its acquisition by Alteryx in 2022. In this interview, Adam shares his expertise and insights on harnessing the potential of data analytics to fuel business innovation.

I'd rather read the transcript of this conversation please!

Intro - 00:00:02: Data Masters is the go-to place for data enthusiasts. We speak with data leaders from around the world about data, analytics and the emerging technologies and techniques data-savvy organizations are tapping into to gain a competitive advantage. Our experts also share their opinions and perspectives about the hyped and overhyped industry trends we may all be geeking out over. Join the Data Masters Podcast with your host Anthony Deighton, Data Products General Manager at Tamr.

Anthony Deighton - 00:00:38: Welcome to another episode of Data Masters. Today's guest is Adam Wilson, Senior Vice President and General Manager for the Alteryx Analytics Cloud. Alteryx unifies analytics, data science, and business process automation in one end-to-end platform to accelerate digital transformation and shape the future of analytic process automation. Away from Alteryx, Adam keeps himself busy as a limited partner at Greylock and Ridge Ventures. He's also a Technology Executive Council member for the CNBC Technology Executive Council and an investor and strategic advisor for Mach5 Software. Welcome, Adam.

Adam Wilson - 00:01:18: Great to be here. Thanks for having me.

Anthony Deighton - 00:01:20: So I thought we would start at the very beginning and we'll talk about the founding story of the startup you founded, Trifacta, and listeners may or may not be aware that there's some overlap between the founding story of Trifacta and of Tamr. But maybe share in your words how you came to be a part of Trifacta and the founding story, Trifacta.

Adam Wilson - 00:01:46: Yeah, it's very interesting. So Trifacta was actually started as a joint research project between a professor at Stanford focused on human computer interaction and visualization and a professor at Berkeley that was focused on databases and distributed computing. So that was Jeff Heer and Joe Hellerstein luminaries in their areas. And they had a shared grad student, Sean Kandel, who was doing work in and around this idea of, how do we enable people who know the data best to do some of this very complex work around data wrangling? And I guess the tie to Tamr was that early on, Joe, who had been a student of Mike Stonebraker along the way, was sitting around the kitchen table talking about how this was such a huge problem. And Joe was doing a research project that was focused on human computer interaction and how do we enable people who know the data best to do a company together focused on this? And as it played out, I think Joe felt very passionately about thinking through how the human in the loop, how do we really think about more of a co-pilot approach to solving the problem? I think Mike at the time, and this is again in the early origins of these companies, was a little more like, how do we get the machines to do more of the work. So there was a little bit of a philosophical difference in approach, although I think probably with the passage of time, elements of both ideas crept into both companies in pretty meaningful ways. But it was really, for Trifacta, those sort of that early formative experience of like, how do we turn this into a user experience problem powered by machine learning that led to Stanford Data Wrangler, which within six months, there were fifty thousand people using it. And that's when our good friends at Excel and Greylock came knocking and said, hey guys, we think there's a real company here, not just an academic prototype. And it was professors take sabbaticals and Sean wrapped up his degree in a hustle and jumped out to do the company for real.

Anthony Deighton - 00:03:42: Yeah, and it was quite a ride. So you guys built an amazing product. I was actually a user of it in its early days. But then about a year and a half ago, Alteryx announced the acquisition of Trifacta. Now you've been working, as I suggested in the introduction, you've been working at Alteryx and really bringing the two ideas, the concept of Alteryx and the concept of Trifacta together. And so maybe talk a little bit about what it means to bring these two companies together, how that's changed what you're doing and what the products are doing and how the two really work together to add a lot of value from a customer's perspective?

Adam Wilson - 00:04:20: Yeah, it was interesting for us along the way. So, over almost eight or nine year journey from the original academic project to bringing on first customers and starting to really finding product market fit and then beginning to scale the business and signing up strategic partnerships like Google who brought Trifacta to market as a Google branded service first time in their history they'd ever done that. There were all these amazing milestones along the way. But what was interesting is, as we were getting into accounts and as we were starting to see meaningful expansion, you know, Alteryx was in every account that we were in. And it was, we didn't necessarily bump into them head to head a lot. And so it was obviously a very successful brand and company that I was very aware of, but they were landing with end users in the line of business. They were taking a lot of work they were typically doing with files and automating that in really meaningful ways, but like going to, you know, people more like business users, accountants, you know, researchers, analysts and helping them understand like, you can now create your own data products and you can do that in a very self-service way. So again, very akin to a lot philosophically, a lot of where Trifacta, you know, had come from, you know, and, but yet we were going in through IT organizations. And so we were often part of data modernization initiatives where people were thinking about what the cloud is going to mean to data and analytics? And how do I contend with increasingly larger and more complex data sets? And how am I going to use machine learning and AI to speed up everything and to allow us to take some of the most complicated things and automate those, you know, because every data set's not a new data set. Odds are somebody has seen something like it before and you can even learn as the user interacts with the data. And so coming through IT with these modernization exercises, Alteryx had really won hearts and minds of end users in the line of business. And so when Mark and I first had a discussion and we got connected through some shared connections in the VC, there were board members that Mark had had, Palo Alto Networks who were involved with Trifacta. They said, you guys should meet up. You know, Mark had just taken the reins at Alteryx and as we sat down, it was really interesting because we were realizing like a lot of what we were trying to do was solving a lot of the same kinds of problems, really, you know, shared a vision for how we could democratize a lot of this work that had historically been the exclusive purview, the highly technical, you know, had been coming at it from very different ways. And so, you know, for Mark, it was like, I want to accelerate everything we're doing with the cloud, you know, and go bigger, faster. And for Trifacta, it was like, hey, we've got this really great cloud solution, but we're a small company, you know, we need awareness and distribution. And so it was a little like, huh, this seems perfect. So pretty quickly, hey, maybe this is interesting and maybe we can combine forces and help each other out in meaningful ways to, you know, over the course of, I don't know, about a six month period or so, you know, becoming one company. And I think, you know, for me personally, and for really for the entire, you know, Trifacta team, for the Trifacta customers, it's been amazing, you know, to now come in and now being part of a larger public company, Trifacta really became the foundation, the platform for Alteryx Analytics Cloud. And so not just for prep and blend, but also for auto machine learning capabilities, auto insights capabilities, location intelligence capabilities, a lot of applications plugging into that platform and leveraging, you know, eight, nine years of hard knocks and experience that we'd earned, you know, in the market. And so a lot of that original thesis, you know, very much played out, you know, it's like, can this accelerate, you know, not just in weeks, months and quarters, but literally in many years, the ability to deliver meaningful and integrated, you know, cloud solutions to the market. And so we've been super busy with that over the last months and have already launched Better Together offerings, you know, based on this. And for us, you know, also just saw, you know, an ability to now do things in a bigger way on a bigger stage with increased investment, more specialization that has been incredibly powerful for us. So it's been a lot of fun. It's been a crazy busy year one, as you can imagine, just all the post integration things that happen when you stitch together two companies, two cultures in a variety of technologies, but it's also been really rewarding. And I think we're headed to our user conference here later this month and are going to be unveiling, you know, a whole set of new ideas and applications that really came out of that collaboration, you know, over the last months.

Anthony Deighton - 00:09:12: That's awesome. So it sounds a lot like this idea of accelerating Alteryx moves in the cloud. That thesis really played out positively. And this idea that we want to empower and I love this word democratize. We want to empower end users to work with data in a very tangible way through the user interface and through these take advantage of the scale and compute that you get on the cloud. Is that a fair way to bring the two together?

Adam Wilson - 00:09:37: Yeah, absolutely. And I think that to some extent, if you think about self service and you think about democratization broadly, a lot of it is about removing frictions and barriers. And clouds are great at that. You know, all of a sudden a whole bunch of the complexity that companies bump into, especially as the data volumes get larger, as the data diversity gets increases, you can leverage a lot of cloud services to abstract, you know, from some of that, or at least inflect that back in a way that the end users don't have to contend with it so much. And so I think that has created a lot of opportunity then to, you know, rethink like, all right, well, you know, where is this work being done? Who's doing this work? And how do we do it most effectively and efficiently? And so most of the organizations I talked to are looking at over the next few years, you know, doing substantive portions of their data and analytics work in the cloud. It's interesting, though, because not all right, meaning not so much most of the customers I talked to are but not all of the work is contemplated to go to the cloud. Many of the larger organizations are pretty committed to hybrid, pretty committed to the idea that like, hey, some of this is going to stay on prem for a variety of reasons. Sometimes it's, you know, related to security. Sometimes it's just related to, you know, certain internal processes and efficiencies that they gain by doing work on prem. So you're seeing this interesting blend where, you know, more aggressive moves the cloud than ever before. At the same time, some organizations retrench, you know, certain types of workloads around doing that work in their own data centers.

Anthony Deighton - 00:11:13: So I want to pull on that thread a little bit. We've talked a little bit about Trifacta and Alteryx. So if you step up a little bit and think about data in the enterprise. So you've talked about one trend, which is the move of data to the cloud. And as you've pointed out rightly so, I think that it's really a hybrid approach, some data on the cloud, some data still on-prem. But if you step back and think about what you see going on in data and analytics today, one trend, what are some of the other trends that you see going on in data and analytics?

Adam Wilson - 00:11:44: Well, I'm surprised we've gotten this far into the conversation without somebody saying ChatGPT or OpenAI. I was at dinner here in San Francisco the other night and the waiter walked up to the table and was just introducing himself and asked what line of work I was in. I said, I work in technology. And he immediately started asking me about ChatGPT. And I was like, okay, if now this is like the dinner conversation I'm having with my waiter in San Francisco, like, you know that this has become pretty pervasive. So, but yeah, no, I mean, I think, you know, to me, you look at, you know, this is very of the moment, but like you start looking at that as a trend and, you know, so what are we doing? Right. We're creating even increasingly user friendly interfaces to do complex things, you know, so not to oversimplify it, but, you know, the idea that, hey, this is going to replace some manual tasks or going to enable a broader set of users to go in and, you know, ask and answer more complex questions. I think that's pretty magical and that's super exciting. I think that people always ask, okay, well, does this replace people? Does it replace jobs? And I'm like, you know, we've seen these trends, more abstraction, more automation, you know, more user friendly experiences, we've seen waves of them over decades. Right. And in each case, you know, it has tended to replace, you know, certain activities, often time intensive activities, often very painful and manual activities, but it's not a replacement for the creativity of the person that ultimately is making the business decisions or asking the questions. And I think, you know, for me, that's always been so core, which is like, you know, context matters, you know, the ability for people to think creatively matters. And, you know, there's a lot of distance between questions and ideas and being able to get at the data that will help welcome uncertainty into your business, cater to long segments in your market, manage and model risk better. And anything we can do to close that gap is beneficial, frees people up, you know, again, implement a lot of the, you know, the business insights, you know, and to take action on those insights. And so I think I just look at it as yet another thing that we will add to our arsenal. And I think that the market is trying to figure out how to do that in a responsible way, because the technology is changing so quickly and there's so much heat on this right now that, you know, I talked to the CTO of a Major Bank last week who just said, we shut it all down because we just don't fully understand it yet. Right now, we're more worried about the risk than the benefit. But he believes deeply that there is benefit. But he's looking for the ecosystem, you know, companies like Tamr and Alteryx and others to like that are trusted, that have been in there doing a lot of this work for them for many, many years to help demystify, help integrate and incorporate and help them adopt this in a responsible way. And so I think the ecosystem, you know, a lot of us that are on the bleeding edge or more avant garde have to like really help a lot of these large organizations, you know, start to think about ways in which they're going to get real benefits from it, but in a way that, you know, still provides the kinds of governance and protection and controls that highly regulated environments like banks and health care companies are going to need.

Anthony Deighton - 00:15:12: So I think that's a really good point. Maybe a slightly different way of saying that is in the application of these technologies that we see their real value and organizations are struggling with figuring out how to apply them. And the other, I think, important insight you're sharing there is the where we see automation and artificial intelligence stepping in is actually removing the worst and most boring work that we do, which I think is very consistent with what the Alteryx Analytics Cloud is doing, what Tamr is doing, what Trifacta started to do, which is like in general, working with data is a complete pain. It's boring. It's repetitive. It's frustrating. It's time consuming. And if we could just not do that and actually focus energy on the interesting problem of, you know, looking at our business and understanding margins or whatever the problem of the moment is, yeah, that actually probably would be better.

Adam Wilson - 00:16:07: Yeah, I mean, we always used to talk about how, you know, at Trifacta, we were very proudly data janitors. You know, I think for us, you know, this was this idea of like, hey, it all starts with the data. This recent trend underscores that massively, right? Like if your data quality is bad, then your machine learning, your AI, your algorithms are worthless. And I think that this does create a burning platform to also really start to think, you know, more comprehensively about how you're building your data products. You know, it starts to be a game of who's got access to the most interesting sets of data and can bring those data sets together as quickly as possible in order to, again, get to unique insights. And, you know, and God forbid that, you know, if you start automating using AI, start automating bad decisions faster based on bad data. Right. Like, I mean, you know, so you just have a disaster just because it happens faster. That's right. So you have to be really careful about it. And, you know, there's an example that I've shared in the past. You know, again, going back to banking, you know, we work pretty closely with an algorithmic trading group and they were like, listen, you know, these are data-driven, data savvy people, but they're not necessarily structured programmers. You know, they're like, I have a voracious appetite for different cuts of data combined in new and interesting ways because. that would birth algorithms that I can trade on and if I'm a little faster, I don't just win a little more of it. Like I win all of it for some period of time until everybody else catches up to what I'm doing. And so this is where you start seeing that like the frontier for competition now becomes, you know, can I find new and interesting data? Can I onboard that new and interesting data? Can I start to like to stitch it together in ways that will create training sets that will birth algorithms or that I can run across, you know, to get these unique insights? And if you can speed that process up, you know, that's not just an operational savings, like do more with less message. That's a top line, you know, beat your competition and win more shares and do better for your customers' message. And that's ultimately really powerful. So I think these trends that we're seeing right now, just, you know, put an exclamation point on the importance of the data strategy, the data platforms and foundations, the being able to do these things at scale and really emphasize these things like data quality more and more.

Anthony Deighton - 00:18:29: Amazing points. And another element of the GPT phenomenon is the chat part of the ChatGPT. And this is one I think doesn't get talked about as much, but the very fact that you might have a waiter in San Francisco or, you know, a random family member talking to about engaging with a large language model in this specific instance through a chat interface means that this idea of artificial intelligence for people to engage with on a daily basis to do things like write poems and you know silly things almost it takes it out of the scientific department and puts it into the general public and that's a little bit of I think where we see these technologies become widely adopted is when they become something that everybody understand and you understand things by playing with them, with that as a backdrop and you're also involved in a number of different venture funds it's an interesting time as they say a euphemism in the VC Space today I'm curious on what you see going on in venture capital is anyone getting funding you know and maybe even in the data and analytics space generally like what kinds of beyond GPTs and Large Language Models what you see going on in the VC Space

Adam Wilson - 00:19:49: Well, what's interesting is that, I don't know, what you read a lot about these days is how there's an extinction event that's happening and nobody can get financed and everyone probably didn't slow down their burn fast enough. We're overvalued and under ARR. So now it's just creating all these negative effects and making it really hard on organizations. And I think there's a lot of truth to a bunch of that, right? So that's not... The one thing that I think is specifically interesting in this area of data and AI is that there is an emerging idea that a whole new class of applications is going to be born out of this. And that if you didn't build from the ground up with these fundamental assumptions of what can be done with Large Language Models and what can be done with some of these new interfaces that maybe some of the legacy CRM applications, SFA applications, will have a hard time bolting that on and being competitive. It's like, what if your HR systems could tell you when people should or likely to leave or could tell you when it's time to think about promoting them and it could recommend skills that they needed to develop and training and curriculum. Can you actually take something that doesn't start with data and analytics from an application layer at its core and with AI at its core and think about every single place where that could make whatever you're doing now better, faster, stronger? And can you layer that into some of these applications that have been around now for 20, 30 even longer periods of time? And so I think there's excitement in the venture community that this is going to bring on another wave of disruption where you can build some meaningful, scalable businesses. And I think that's the stuff that VCs always get super excited about is these moments in time where you just see these secular megatrends or these market shifts that occur. And it's not just about feeding the incumbents all the time. So I think in that sense, it's a really exciting time. That being said, and there's plenty of capital out there. Especially with what's been going on in the markets right now, people are looking for places to put money to get return. I think the hard part, though, is that everyone's having to be much more grounded and realistic about what it's going to mean to raise around what valuations are going to be. And so I think adjusting expectations there is going to be crucial. You're seeing a lot of organizations, too, that private equity becomes a viable option if they're out there because they're going to need to recap and reset, or they're going to need to combine forces and get stitched together with other assets in order to create something more meaningful and more substantive. I think the days of thin slicing the analytics stack, which was, again, I think also a VC fed thing, where it was like, hey, in order to do anything end to end, you're going to need 15 different products. And one of these products all said, yeah, you need one of me and then one of them. I think those days are gone as well, where organizations are like, that's just too complicated. The tool chain exploded too much, too fast. And there's too much headache and coordinating across all of that. And so you're starting to see platforms become more prominent and organizations that have, you know, some critical mass where they can make long-term bets and then start to layer in other capabilities over time become more prevalent. And so, I don't know, those are a few of the trends that I guess I'm seeing. So I don't know if that matters, you know, what you're seeing as well.

Anthony Deighton - 00:23:43: It definitely does. So this move to startups needing to be more thoughtful about how they generate revenue and how they find customers and how they actually bring these technologies to market has also maps nicely to the need for these technologies to solve real problems. Just because you've built artificial intelligence, the machine learning model in itself is interesting, to use your example, if you apply it to the domain of your example, human resources, now it starts to get interesting because now I don't care that it actually uses machine learning under the cover. I understand that it solves a problem I have with retention or turnover. Great, that's a problem I'm willing to pay for. And at its core, all of these things are really about taking advantage of the data that the enterprise is generating and having better meaningful and scalable ways of being able to use that data to make better decisions and be more effective at the job. So probably this is a return to some version of normalcy which itself will be a useful filter and accelerant to real innovation.

Adam Wilson - 00:24:42: Yeah, I agree. I think that's into your point about, you think about all the interesting applications around fraud, detection and insurance adjustment and cancer research and claims adjudication. You know, it's just like the richness of what's going to happen, I think, in the app space, but specifically in the either vertical or functional domains. I think, you know, there's going to be a whole bunch of ready problems where the degree of automation is going to just increase massively, right? And it's going to be very targeted, it's going to be very specialized, and it's going to be incredibly powerful. And I think that, again, I think that's super exciting. I think it's exciting for everybody broadly, but I think coming back to your original question, I think that's particularly compelling in the venture space right now. And you're starting to see a lot of early stage investments in companies that are taking that approach. It is not lost on me as well that for those of us who are, you know, broadly in the data space, we're arms dealers to that segment as well, right? Because it's like every single one of those products, those applications, those vertical initiatives all need to be able to get the data together quickly. They need to be able to make sure the data is appropriately cleaned. But then also, it's like even generating some of the insights. Like, you know, we get to a point now in one of the products that I'm very involved in at Alteryx is an auto insights product. And the whole thesis is like, why can't the datasets tell you what's most interesting about them? Like, why do you have to wander the data set or why do you have to stumble across the dashboard that might have the answer to a question that you have? Like, why can't just this be continually happening as new data comes in? The data sets are surfacing what's where the anomalies, where the trends, they're already doing all the dimensional analysis for you and coming back and saying, these are the things we think you ought to pay attention to. Right. Or this is where you should probably dig in. You know, it's like that automation, moves beyond static reports and dashboards and moves beyond what you've historically seen and really empowers again. The end user now gives them a running start and gives them a place to focus their exploration. And I just think to me, that's the really interesting stuff that you're starting to see happen that I think is going to be transformational to what's going on in analytics.

Anthony Deighton - 00:27:06: Heavily dependent on machine learning and the cloud. I mean, ultimately, those are the enabling technologies that allow these differentiated user experiences to blossom. Well, look, Adam, a real pleasure. I appreciate you sharing the founding story of Trifacta and the journey to Alteryx and the interesting success that we've been having together to really help to transform the data and analytics space and interesting commentary on the venture community and what's going to take us into the future. So thanks for your time today.

Adam Wilson - 00:27:38: Thanks for having me. It's always good to reconnect and certainly a lot of fun.

Outro - 00:27:43: Data Masters is brought to you by Tamr, the leader in data products. Visit tamr.com to learn how Tamr helps data teams quickly improve the quality and accuracy of their customer and company data. Be sure to click subscribe so you don't miss any future episodes. On behalf of the team here at Tamr, thanks for listening.

Suscribe to the Data Masters podcast series

Apple Podcasts
Google Podcasts