In this episode, Dana Gardner is joined by Zach Maybury, Chief Technology Officer at DraftKings to discuss the intricate balance of market agility and regulatory compliance for the sports entertainment company. They explore DraftKings' cloud-native roots, hybrid cloud strategies, real-time data ingestion, and staying agile in a rapidly evolving industry.
In this episode, Dana Gardner is joined by Zach Maybury, Chief Technology Officer at DraftKings to discuss the intricate balance of market agility and regulatory compliance for the sports entertainment company. They explore DraftKings' cloud-native roots, hybrid cloud strategies, real-time data ingestion, and staying agile in a rapidly evolving industry.
[00:00:00] Producer: Hello, and welcome to the Data Cloud Podcast. Today's episode features an interview with Zach Maybury, Chief Technology Officer at DraftKings. Hosted by Dana Gardner, Principal Analyst at Interarbor Solutions. They discuss the intricate balance of market agility and regulatory compliance for the sports entertainment company.
[00:00:23] Producer: Dana and Zach also explore DraftKings cloud native roots, hybrid cloud strategies, and staying agile in a rapidly evolving industry. So please enjoy this interview between Zach Maybury and your host, Dana Gardner.
[00:00:35] Dana Gardner: Welcome to the Data Cloud podcast. Zach, we're delighted to have you with us.
[00:00:39] Zach Maybury: Delighted to be here, Dana. Thanks for having me.
[00:00:41] Dana Gardner: You bet. The twin requirements of market agility and strict regulatory compliance create an especially complicated challenge for data gathering, analytics and reporting management at DraftKings. Describe DraftKings, its still evolving legal environment and the need to manage compliance across such a large, dynamic and diverse landscape as 30 separate state jurisdictions in the US alone.
[00:01:08] Zach Maybury: Absolutely. Well, DraftKings is a sports, entertainment and gaming company and we are industry leading, have their top Eilers app for our sportsbook or casino. We have daily fantasy products, a lottery as well through Jackpocket and are always looking to expand into new product verticals. And as you can imagine, that creates an incredibly complicated data environment and makes working with fantastic partners critical.
[00:01:34] Zach Maybury: It makes having world-class staff critical, and you mentioned the regulatory requirements across all of our products. This is a critical component for us to get right. So we start by working with regulators to understand what those requirements are and how best to achieve the customer outcomes that we want, but working within the regulatory framework.
[00:01:55] Zach Maybury: That varies state by state, which means that there's a lot of interesting technical challenges, complexity, configuration that we have to build to navigate all of that. So across all of our products, each one is a little bit different and we have to work across each of the regulations to bring product to market.
[00:02:15] Zach Maybury: So it creates that incredibly complicated, you know, technical ecosystem, which is one of the things that the engineering team loves is taking that complexity head on and figuring out how to best deliver for our customers as the world doesn't stop moving. At DraftKings, we like to say we move at the speed of sports, and so navigating the regulations to be able to deliver products that customers love is absolutely the mission we take on every day.
[00:02:42] Dana Gardner: And Zach as CTO, that can mean different things at different places. Tell us about your role as CTO there, your background and how you got into this business.
[00:02:53] Zach Maybury: So my role, of course, it's technical being the Chief Technology Officer, but it really starts with team. My mission is not to dictate.
[00:03:05] Zach Maybury: The technology and the strategy that the team does, that's one of my responsibilities, but my mission is much larger than that. My mission is really, do you have the right team? Are they empowered and can they deliver business like ours? Over 2000 people in engineering. That's a huge enterprise and it requires everyone working together, being highly mission oriented and aligned to deliver all across all of these products and to keep things coming, remain agile for things to not slow down.
[00:03:36] Zach Maybury: So my mission is actually more about culture, people, and team empowerment. And then everything else flows from that. If you get those things right, teams feel empowered, they build great technology, they tackle problems head on. They don't shy away from the complexity. So my job every day is meeting with teams and asking those hard questions, you know, are we delivering for the customer, foremost?
[00:04:02] Zach Maybury: Does the team have what they need? Are they empowered to deliver against that mission? And then from there, I know what are the roadblocks? What are the impediments? And that's really my job, is to figure out how to remove those, how to help the team work around them or work past them to deliver the best apps on the planet.
[00:04:22] Dana Gardner: And of course, online sportsbook is a fairly new business and is rapidly changing, and I should think that the requirements are of a nature where you probably couldn't have done this without a digital first culture and team. Tell us how data optimization plays a role in how it's making possible something that probably wasn't possible, not that long ago.
[00:04:47] Zach Maybury: Yeah, so our roots are cloud native. Back when we were just the Daily Fantasy company pre PASPA in 2018, we were a hundred percent on AWS cloud native and really looking to innovate with the best native services that are available on Amazon.
[00:05:07] Zach Maybury: And so those routes, I think are very important. Combine that with our analytical, data oriented structure, it sets us up really well for the new industries that have emerged in online sportsbook, in iGaming, as well as now the lottery space. So each one of those brings its own set of regulatory complexities, data complexities, customer personas are different across all of those.
[00:05:33] Zach Maybury: And so, I think we're incredibly fortunate that we grew up native in the cloud as a company, as new product verticals came along. We were thinking in a cloud first way. We're thinking about how are we gonna ingest new data volumes, especially the hybrid cloud nature of the regulated sports betting landscape.
[00:05:56] Zach Maybury: Being cloud first, thinking about scale and complexity, I think made it easier for us to sort of transition, I'll say from cloud native to hybrid. For us it was just another step in complexity across our cloud ecosystem.
[00:06:13] Dana Gardner: Yeah. Let's drill into that a bit, if you don't mind, Zach. Explain why you need to have an on-premises capability and hence the hybrid nature of your architecture when it comes to these jurisdictions.
[00:06:24] Zach Maybury: Yeah, so one of the key components for online sportsbook is that the wager itself is accepted within the state that the customer is placing the wager on. So in the cloud, the services could be running potentially anywhere. We, of course, have a little bit of control of that based on what region we're requesting the capacity from.
[00:06:44] Zach Maybury: But could move across data centers within a region, and that creates a problem or a complexity when it comes to navigating in-state requirements. And just to underscore, each state sets its own requirements. So the requirements in New Jersey are different from the requirements in Virginia and requires us to take a different infrastructure approach in each of those states.
[00:07:08] Zach Maybury: In some states, we operate in a co-location or even in a data center. In other states, we work out of a local zone, and in some states we even actually work right out of the primary cloud region. So, each of these states has workloads, has databases, some synchronous, some asynchronous, and all of that needs to come together and make its way ultimately back home in a very low latency way to our data lake warehouses.
[00:07:35] Zach Maybury: All of the reporting has to tie in together and has to balance out. And so this, you know, jump in complexity is really what the team spends its time tackling.
[00:07:48] Dana Gardner: Yeah, so that makes you a poster child of managing and coming up with the best operating approach and best practices for having to have data locally housed.
[00:08:00] Dana Gardner: And even the transactions take place on-premises in certain places. But you also wanna get the best of that cloud first centralization. So tell us how you've been able to manage the best of centralization and the best of distributed operations. 'cause I think a lot of people are gonna be having to grapple with that as time goes on.
[00:08:18] Zach Maybury: Yeah, it absolutely starts with a customer focused approach first. That quickly leads into architectural discussions and really the things we're focusing on there are the customer success requirements that we're looking to achieve. So we focus a lot on end user latency. Efficiency is important. We have seen and we do believe that we get more efficient the more cloud native we become from a cost perspective.
[00:08:47] Zach Maybury: But really the primary thing that we're focusing on is those customer outcomes. And so we start there and we try to do as much work as possible architecturally, as close as co-located as possible. And a lot of that key data is for us cloud native. And so we try to do as much compute as we can in the cloud.
[00:09:11] Zach Maybury: From there though, when we reach either a regulatory requirement or there's sort of a end user benefit to be more in our edge network or in our in-state network, this is where we start to move the workloads to the state. So that placement is obviously key there from a regulatory standpoint, but there are other key tasks that we look to accomplish in-state.
[00:09:37] Zach Maybury: Particularly things around disaster recovery, there's some reporting. Our casino games also run in-state, and so that's a really interesting thing to source all of the assets, the game engines and the math that power all of our casino games in-state as well. So for us, it's not as simple, unfortunately, or maybe fortunately, the engineers, I think love the complexity and in working at the scale that we work at.
[00:10:02] Zach Maybury: But, it's not as easy as just sort of spinning up an environment, spinning up your workloads and having everything, you know, nice and co-located. I can give you a specific example that I think is really interesting. We recently have started to use more direct connections between our in-state presence and our cloud workload.
[00:10:23] Zach Maybury: So previously, to communicate between in-state and our cloud data environments, we would mostly just use the public internet. And the public internet works great. It's a, you know, best effort network. It powers the vast majority of what everyone does every day. So this is not a knock on the public internet.
[00:10:40] Zach Maybury: However, in the public space things do go wrong. The networks need to be rebalanced. You don't always get the most optimal route. It's a best effort design, a best effort, resilient design. And so we have actually taken the step to force wrecked backbone connections between our data centers and our cloud providers, Amazon and GCP.
[00:11:02] Zach Maybury: And we saw great latency improvements here for our end customer. Sometimes up to 75 milliseconds. Now 75 milliseconds, you may be going, that's not a lot, Zach. It's a lot because in any one transaction, a lot of things need to happen to, ultimately, to place a bet. We need to verify that you're located where you say you're located.
[00:11:24] Zach Maybury: We have to verify that you have the funds that you're intending to use. We have to verify that there aren't any business reasons why we shouldn't take the bet. Are there risks for fraud concerns that we have? You start to add these up. All of these checks, they need to happen in less than a second.
[00:11:42] Zach Maybury: Otherwise, the user starts to really perceive this latency. So when you can save 75 milliseconds just in the wire to wire communication of your data, it actually means a lot to the end user. And it also leaves a lot of pressure on the rest of your ecosystem in terms of the latency that you can tolerate.
[00:12:02] Dana Gardner: Yeah. And in optimizing for the latency and for all those different types of transactions going in and out of all sorts of different providers and third parties, have you relied on your cloud providers to manage a lot of that, or have you had to do it in-house? To what extent is this a shared endeavor between you and your providers?
[00:12:21] Dana Gardner: And I'm especially interested in that, you know, when it comes to the ability to use, maybe, content delivery networks or other tools.
[00:12:30] Zach Maybury: So we depend and rely on fantastic partners for much of our solution. We also own a lot in-house, so I think the real answer is it's both. There's a lot that we need to own in the first order way we feel to deliver the best product in the industry.
[00:12:48] Zach Maybury: However, where we have content delivery networks, Amazon native services, the ecosystem there around private link and local zones, these are critical to our operations as well. And I would say we've learned a lot over the years. You know, when we started in New Jersey in 2018, this was a regulatory requirement, but we ran in a space, in our partner’s location at resorts in Atlantic City.
[00:13:20] Zach Maybury: Atlantic City, it's right on the coast. It's not where you would build a data center if you could choose anywhere on the map. And so with that comes real world challenges. Everything from weather and power to just the physical location itself and your ability to add more headroom or scale.
[00:13:36] Zach Maybury: All the reasons why you wanna not have to think about any of this is, you know, why you wanna hyperscaler, it's why you wanna work with Amazon. 'cause you don't wanna think about what happens when I run outta space in my rack, or do I have a third source of network egress, or do I have a backup generator and did I remember to fuel it?
[00:13:54] Zach Maybury: So over the years, I think we've evolved a lot from our early days in New Jersey, adding new states, each state allowed us to build more configuration and flexibility into the platform. As well as learn a bunch of key lessons on where we wanna be spending our time or where it makes sense to bring in a partner to help us.
[00:14:16] Zach Maybury: And moving to local zones, I think is a great example of that. We saw great improvements both in speed and latency, as well as just cost in general. Moving from our own self-built, purpose-built, managed racks that were sort of in co-location settings. Moving to a managed service private cloud model with AWS local zones is just sort of with the flip of a switch.
[00:14:41] Zach Maybury: Our efficiency got better. Our stress went down and it was a big win all around. So we have an extremely high bar for our partners. So, you know, when we work with Snowflake, when we work with AWS, you've met a high bar in order to sort of be in the room with us. But once we're all together, we're one team trying to deliver the world's best sportsbook product.
[00:15:05] Dana Gardner: Right. And you mentioned reducing risk and fraud and security. I mean, for most organizations that means keeping the bad people out, but in your case, you need to make sure that the inappropriate people don't come in. And so it seems to me that there's a lot of data management in near real time again to make sure that the people who are partaking in these services are the right people to be doing that.
[00:15:27] Dana Gardner: So how does that factor and how does, perhaps, the ability to democratize data science across your organization give more people the opportunity to help you keep the right customers in the right roles?
[00:15:42] Zach Maybury: Yeah. The democratization of data and data science across the org is critical for so many reasons.
[00:15:50] Zach Maybury: There's so many customer-facing applications, things that we're interested around personalization and content delivery. Sort of, the right user, the right merchandise, the right time is critical. But as you mentioned, I think the vast majority of the business's focus is actually on things that customers never see.
[00:16:14] Zach Maybury: It is on what does it look like to comply and to, honestly, to even exceed the expectations on things like anti-money laundering or responsible gaming. These are absolutely critical for our business to be successful. We have to get those things right and we have to get them right every time. On the flip side, the vast majority of customers, the vast majority of traffic doesn't fall into those pockets.
[00:16:37] Zach Maybury: It's someone who is, you know, trying to place a wager, to play a game. Draw a fantasy contest, buy a lottery ticket. They're just going about their life, and DraftKings is providing some joy into their experience. And so, the last thing we want is for those customers to kind of get unnecessarily caught up in the friction that we are intentionally creating for bad actors, let's call them.
[00:17:02] Zach Maybury: This doesn't even get into, you know, threat vectors and actors that are attempting to create denial of service attacks or steal data or other things like that. So there's the criticality of understanding your data, understanding your customer, understanding how those things come together, and using that, democratizing that so that teams can make the right decision virtually every time, virtually instantaneously.
[00:17:31] Zach Maybury: That's where the magic really happens. Going back to that example, talking about the latency in the stack, we wanna get their answer right every time, but if it takes us 10 minutes to do it, that moment has already passed for someone who wants to place a sports bet or wants to make a deposit and their money is hung up, awaiting manual review.
[00:17:55] Zach Maybury: So automation is absolutely critical. And I said 10 minutes, nevermind 10 minutes, like, a hundred milliseconds might be too long for some of these decisions. So the amount of engineering work that goes on behind the scenes to build advanced, near realtime and realtime models to detect bad actors with virtually perfect accuracy impeding our customers is really mind boggling when you look at how much time, effort, and care the team spend on building those solutions.
[00:18:28] Dana Gardner: Yeah, it's very impressive now that I've learned more about, you know, what your requirements are and the challenges. And so, we've talked about the requirements for jurisdictional compliance.
[00:18:39] Dana Gardner: We've talked about security and risk management. We've talked about topology and taking the best of decentralization and centralized cloud native computing. But you also are in, and I hesitate to call you a startup because you're very impressive in how mature you are technically, but you're still evolving in how to best provide services and engagement and entertainment to your customers.
[00:19:02] Dana Gardner: So when you bring in novel, new services, maybe you want to try them out on an AB testing basis. How is the architecture and your vendors and suppliers supporting you and the ability to be agile when it comes to adding new services and then getting the reporting back on how well they received?
[00:19:20] Zach Maybury: Yeah, you said the key words there. You know, are we a startup or not? I think regardless of how we're classified, what's key is the values of what being a startup looks like. Now, we don't want sort of churn or uneasiness, but what we're absolutely going for is a rapid agility and unlocking maximal value as quickly as possible.
[00:19:42] Zach Maybury: And so those aspects of our, I'll call it our startup culture have persisted and I think really drive, you know, our world class engineering team to continued new heights. I'm constantly floored by when myself, our Chief Product Officer, Paul Liberman, one of our co-founders, when we step back and we look at just the sheer amount of what the teams have accomplished.
[00:20:07] Zach Maybury: Like for example, going into this year's NFL season, it's honestly, it floors us how much the teams are able to do well, continuing to run this massive enterprise day in and day out at incredible levels of resiliency and reliability. So I think this balance of not losing your speed and agility is really key.
[00:20:30] Zach Maybury: Making the right architectural choices, like we are heavy Kafka and asynchronous flow consumers, and at our scale, I think this is actually critical to allow us to sort of continue to scale horizontally and build a very resilient low latency, but also flexible ecosystem. Choices like that allow us to plug in new things more quickly.
[00:20:56] Zach Maybury: So when we have a merger and acquisition, for example, and a lot of places, it may be really hard to get that team into your mainline engineering systems, your security or your permissioning, just even your accounts in general. We focus a lot on how do we make those links more seamless, not only to unlock value from those mergers and acquisitions faster, but also because we not only think we know from experience that we wanna launch new products as well, and that makes launching new things faster, ingesting new sources of data, being able to convert that data into our kind of common paradigms, and then be able to turn that into all the things that the business needs around reporting and financials, customer understanding, cross-sell, marketing, and everything that goes into that.
[00:21:53] Dana Gardner: Now, Zach, it's of course the job of a Chief Technology Officer to bring the right solutions to bear. But, you know, the fact of life is as you mature from a startup into a more mainstream and a well-known business, profitability is another important aspect. And so when it comes to total cost of ownership and return on investment, once you've engineered your solutions, what approaches have you had to cut those costs over time and make that technology a part of the business. A slice of the top line work towards your bottom line. So how do you make this cost effective? Do you use what we used to call AI Ops? Can you use the technology to run your systems better or work more closely with your suppliers too?
[00:22:39] Zach Maybury: Yeah. I think one thing that is really helpful is when you make these investments, the next one, the next investment is a lot cheaper. Or maybe sometimes you even get it for free. So sometimes maybe the biggest barrier to starting is the starting itself. When we are building new things, it is a delicate balance between running as quickly as possible to get that thing built into market, which we do very well.
[00:23:11] Zach Maybury: But also along the way, making sure that there's enough best practice embodied in that work that it's not gonna be really difficult later to do the right thing. Ideally, you're doing the right thing from the beginning, so it's much harder to rewrite your schema, all of your stored procedures and sort of change your data model.
[00:23:35] Zach Maybury: After you've achieved some level of scale, it's gonna be way more expensive to do that. It may require downtime. It may not even be possible depending on what other things have been built around that piece of infrastructure, if it's load bearing. So we try to get as much right as possible without slowing down the to market agility.
[00:23:57] Zach Maybury: But to your question about efficiency, I think this is something that we've been focused a lot on the last 18 to 24 months. And the key with efficiency is how do you unlock the low hanging fruit and the efficiencies that you should get anyways, that are just wasteful. These are things that they're not really helping the end user experience.
[00:24:19] Zach Maybury: They're not really making the developer's life better. They're just sort of excess. So we focused a lot on that. What we haven't focused on intentionally is getting razor close to the perfect efficiency number. And I think that would be the advice I would give to anyone is for business like ours, and honestly, I think this is true to any business.
[00:24:42] Zach Maybury: The cost of customer harm is incalculable. And whatever you're gonna save from getting really close to that efficiency point is not worth it. For that potential harm that could come to the customer. So we try to fit kind of right in that channel of smart spend, not perfect spend and not wasteful spend.
[00:25:02] Zach Maybury: And I think what we found is in hitting that channel, it's been actually fantastic working with our partners at AWS and Snowflake because we have inherently a lot of spend that could be smarter. And so really we've been focusing there, a lot of tooling. AI has been great for this. So sometimes the starting point of improvement is knowing where your spend is.
[00:25:28] Zach Maybury: And as you can probably imagine, tagging hygiene. It's one of those things that it's not the most fun sounding thing, but it is absolutely critical if you wanna make a dent in your spend and improve your efficiency is knowing what's actually driving the cost. How is it changing over time?
[00:25:48] Zach Maybury: I think that's where AI has been really helpful for us. Throwing a lot of data, building some custom models, and then basically asking questions about how is this evolving over time? Is this reasonable? We've also actually used it recently in some of our dynamic scaling efforts.
[00:26:08] Zach Maybury: So we do a lot of scheduled based scaling around kind of predictable traffic patterns. So as sporting events come online, customer engagement ramps as comes to the middle of the night, or as there are less sporting events on customer engagement goes down, that works pretty well, but is kind of a blunt instrument if you're saying, okay, it's a NFL Sunday at 9:00 AM, let's scale up all the hardware.
[00:26:28] Zach Maybury: Well, between 9:00 AM and 1:00 PM Eastern, when the games kick off, you actually have a lot of waste. Once 1:00 PM comes around, you're really efficient, but once the Sunday night game ends at midnight. All of a sudden you're inefficient again.
[00:26:45] Zach Maybury: And so we've actually built both in understanding the campaigns that we're running on the marketing side, working with our CRM ops and marketing teams, as well as just looking at how the user users are behaving live on site. We've been able to build some really interesting predictive models that prescale sort of just in time for user demand, and that's really how we've invested to hit the sweet spot of delivering for the customer. Always delivering for the customer, but not getting too close to that razor's edge.
[00:27:16] Dana Gardner: Sure. It sounds like the consumption based model that a cloud native organization would be very familiar with helps and also a lot of the tooling and data that the cloud providers themselves deliver to you to help with that automation and intelligence are very important.
[00:27:33] Zach Maybury: Yeah, for sure. I think that the connectivity to the native services, the observability and reporting that comes alongside of that, the enhanced security, seamless data access, you know, really just even understanding the optimization, the performance optimization, the hardware refreshes all the things that are going on under the hood to make the ecosystem itself run more efficiently means that, you know, we get that benefit, like a rising tide to our efficiency, which has been great for DraftKings as we've scaled.
[00:28:03] Dana Gardner: Sure. So you mentioned earlier that an important relationship is between yourselves and your providers. How about when providers are in a good relationship?
[00:28:12] Dana Gardner: So you mentioned Snowflake and AWS. Is there some way that they work together or are open to innovation among them, between themselves that comes back to help you manage your costs and increase your innovation?
[00:28:27] Zach Maybury: Yeah. The partners that we work best with, it's like they're on our team and it's no different than a DraftKings badge employee versus an Amazon or Snowflake employee.
[00:28:37] Zach Maybury: That all bleeds away and we're all focused on the end customer delivery. The problem that we're trying to solve and strong partnership between our partners, I think, is absolutely key. So you know, we're consuming Amazon native services. We're huge Aurora customers, Mod DB, huge S3 customers being able to natively quickly and efficiently leverage those services.
[00:29:07] Zach Maybury: In our Snowflake ecosystem, our data lake is paramount to our success. So strong relationships between our partners just means that the team is working more efficiently. It means that we can deliver more, deliver faster, and honestly, we can move to get back to the problem we're actually trying to solve, which is whatever feature improvement or a new product that we're trying to bring to market to our customers.
[00:29:39] Dana Gardner: Alright. Last technical question before we start to close out, Zach. I appreciate your patience, but it's been very, very interesting all along. When it comes to ingesting data, because you've done mergers and acquisitions, because you need third party data to manage your marketing and understanding your customers and delivering new experiences, how has the ingesting and management of third party data benefited from one, your internal activities, but also again through your partner topologies and their own relationships with the data?
[00:30:10] Zach Maybury: Yeah, we have always been strongly aligned to efficient third party data ingestion. And this comes from our roots as a fantasy sports company. No one would wanna play a fantasy product if they picked a lineup, and then maybe a couple days later we told them whether they won or not. Customers want to see and breathe and sweat the live experience.
[00:30:34] Zach Maybury: They want the joy of feeling like they're at the top of the leaderboard. Maybe they don't want the agony of falling out of the top of the leaderboard, but they feel that as sort of an innate part of the experience and the only way to accomplish that is ingesting the raw sports events, data streams from various providers as as we offer many sports, each sport has its own way of producing the data.
[00:31:03] Zach Maybury: Its own providers that they work with, their own partners. So in the core of our DNA is flexible, real time ingestion engines, and as we've moved to online sports betting, even casino and lottery. Maybe any product use case has the need for being able to quickly and accurately ingest third party data. The other thing I'd say is, we've gotten bigger as an enterprise.
[00:31:33] Zach Maybury: A lot of our first party data is coming from many, many different sources. So if we think about the data that is produced on the end user application for what a user is clicking on their flow through the application, what emails they're opening and receiving, and you just kind of go on and on, where users are engaging their time, what are they actually converting on, what offers are they engaging with?
[00:32:00] Zach Maybury: All of this data is being created by a number of different development product and operations teams and needs to be ingested, brought in in a canonical way and made sense of to understand the customer profile, to then enable further business operations. Either enhance our marketing that are trading address a fraud or an AML concern address, a responsible gaming incident.
[00:32:28] Zach Maybury: And so in so many ways, I think our strong approach to data ingestion is so much more than just, Hey, what third parties can we quickly plug in and how can we make sense of them in our ecosystem? It even underpins the agility with which our teams in our product verticals, sort of at the edges of closest to the customer, how quickly can they innovate in a lot of ways depends on how well we can ingest that data, understand it, and then turn it into even greater enhanced user experiences.
[00:33:03] Dana Gardner: Well, super. Let's go back to the beginning of our discussion and talk a little bit about your team. You emphasized how important that is clearly running DraftKings as a team sport, technologically.
[00:33:15] Dana Gardner: What is it about putting that team together? Was there anything sort of counterintuitive that you discovered over time, but putting the right people in place to make your achievements possible?
[00:33:26] Zach Maybury: Yeah, I wouldn't say it's counterintuitive, but I think it's hard in practice. You need world-class, brilliant people who are bought in on the mission, are aligned on what they're trying to accomplish.
[00:33:43] Zach Maybury: Are empowered with the tools, the funding, the capabilities that they need to deliver on their ask. And then comes the easy part. You just get out of their way and let them get to work. So it is truly that simple and I'm not the first one to say it, but doing that in practice is very hard.
[00:34:04] Zach Maybury: The hardest thing for me, I'd say, being at DraftKings for almost 10 years in a lot of different roles, a lot of them incredibly technical in nature. The hardest thing for me actually is letting go of the wheel and knowing that by design I've made sure that I have people who are smarter, better positioned and better equipped to deliver on those challenges.
[00:34:26] Zach Maybury: And really what does it mean to empower them to deliver. The other thing I would say that maybe anyone can build towards is a data-driven, forward thinking culture. So we have a lot of people who are not as concerned about what we've done in the past as they are about what we're going to do next.
[00:34:48] Zach Maybury: And I think that is more of a cultural thing and a mindset thing that you can drive into your team. One other thing that we've done for culture that I think is a huge advantage for us is we built a very diverse team in every way you could imagine defining that word. Our headquarters is in Boston.
[00:35:09] Zach Maybury: More than half of our staff, just about half of our staff is across the globe. So we have a huge, very critical, successful offices in in Sofia, Bulgaria and Plovdiv, Bulgaria and office in Tel Aviv, Israel, a big office that we're expanding in London as well as an office in Dublin, Ireland.
[00:35:33] Zach Maybury: And you combine this with our North America DNA, from our old Daily Fantasy sports days, and I think what I'm blessed to lead is a globally diverse team that is empowered, that is hungry to raise the bar and conquer this gaming industry.
[00:35:49] Dana Gardner: Well, thank you so much to our latest Data Cloud Podcast guest, Zach Maybury, Chief Technology Officer at DraftKings based in Boston.
[00:35:57] Dana Gardner: Really appreciate the in-depth look at what you're doing. It is really mind boggling, frankly. So congratulations on that. We really appreciate your sharing your thoughts, experience, and expertise with us all.
[00:36:09] Zach Maybury: Thanks, Dana, for having me.
[00:36:11] Producer: Calling all developers, business leaders, IT execs, and data scientists. Snowflake World Tour is your chance to learn and network. Discover how Snowflake's AI Data Cloud can transform your career and company. Experience the future. Join us on tour. Join the Snowflake World tour to experience the future of the AI Data Cloud with Snowflake. Hear from experts, engage in breakout sessions, and network with peers. Transform your business and career with Snowflake. Register today for one of our 23 stops worldwide at snowflake.com/world-tour.