The Power of Data: Advantages of Data Lakes for Innovation and Impact

[00:04:05] And I think, Altvia is better positioned in a lot of ways than ZipRecruiter was because even though ZIP was very sort of customer-first data-driven strategy, Altvia has this like superpower where we’re not just building the car, but we’re actually along for the road trip too. We’re like saddling up with our customers and we get really into how they do their business, how they think about the world and I think that gives us a unique opportunity to have insights that other companies wouldn’t. 

[00:04:39] So part of the stress, part of the stress and of things that we’re trying to do right now is just get ourselves positioned so that we can try a whole bunch of stuff really quickly and figure out what has the highest impact and then do more of it. It feels good to be back in that sort of startup environment.

[00:04:58] Jeff Williams: I’m gonna rule out that I have to pull this tape out for you at some point because I’ll confess something, which is that from time to time, It feels, and this is the thing for me that perhaps I’ve made up, but it’s like maybe I’m pushing the limits of trying to, move quick on this thing.

[00:05:19] Or perhaps it’s just that sometimes I can be a bit of a hacker and that you, and I know that we’ve talked about that before, but I love hearing that formally out of you because it makes me feel like maybe I have something to rewind and be like you said on the podcast that we gotta move fast and try this, and that’s what I’m doing.

[00:05:38] Jef Rice: Yeah. We try to support all of that. We’re finding the best ideas come from getting in there and getting dirty and working sort of side by side with customers in the trench to figure out what has the highest impact. And you have to be able to do that with a safety net.

[00:05:53] I think part of what we’re doing when we set up our data strategy for our customers is creating an environment that lets them experiment and discover where the hidden treasures are and all of that. But also part of that discovery comes from the fact that you’re gonna fail a lot in figuring it out.

[00:06:11] And if it’s expensive to fail, you’re gonna be reluctant to even try. ’cause I don’t want to do this if it’s gonna take a month and then it’s not gonna work out. But if you could do it in a day and see if it has value. Then you’re okay with the fact that four out of five things are not gonna pan out.

[00:06:28] You can move on to the next thing really quickly. And I think the phase of technology that we’re in right now with AI emerging is in sync with that idea. ’cause nobody knows what AI’s gonna do, we just know that it’s exciting and it’s gonna move really quickly and evolve really quickly.

[00:06:45] And if we have to pick a winner today, we’re probably gonna be wrong. But if we can experiment with certain things and try certain things out in a way that doesn’t take too much time or cost too much money, then ultimately we’re gonna be more successful down the road.

[00:07:00] Jeff Williams: Love it. A lot in there. We’ll come back for some of those things. Want to stop for a second and just I think the theme of this conversation is about data strategy and things like that.

[00:07:11] Jef Rice: Yep.

[00:07:11] Jeff Williams: So I wanna actually what is it that excites you about data? Just why this? Why are you doing what you’re doing? Why data?

[00:07:21] Jef Rice: Aside from the fact that I feel like I’m marginally decent at it?  I think I started my career off on the sort of front-end side of things as a developer. And I enjoy that. I went through a CRM phase, which feels like a teenage goth sort of phase, sometimes like data’s really where it’s at.

[00:07:42] I think data’s super powerful. I’ve seen a couple of different companies that took it. Like I truly believe that the reason ZipRecruiter went from when I joined them, we were at like 150 million in revenue and when I left we were well over a billion and had IPOed and everything else. And it’s because we harnessed the power of the data.

[00:08:01] To make decisions and everybody talks about it, but we truly didn’t do anything unless we could track it and measure it and prove that it was working and we were as fast to throw away an idea as we were to implement an idea. And sometimes I was telling somebody the other day about this we would test stuff by putting a button on a page and the button didn’t do anything, and we were just tracking to see if we could get people to click the button at all. Knowing that we were like, potentially hurting the experience for the people who did click the button and expected it to do something.

[00:08:36] Jeff Williams: Buttons never work.

[00:08:38] Jef Rice: Yeah, seriously. But when you’re doing a million unique visitors a day and you 10,000 of ’em get a bad button, it’s not the end of the world. We had a scale that we could deal with from that standpoint, but the real takeaway is Data, unlocks that capability to be able to say Hey, we’re willing to try anything.

[00:08:57] The best ideas are gonna come from places that you don’t expect. And in a lot of cases and you, I mean you’re a walking, breathing example of this is like you go into a scenario with a theory, and then you pivot on that when the data informs you that thing you were thinking isn’t right.

[00:09:14] But hey, there’s this other thing, and you might end up at the other end of the tunnel with a completely different conclusion than you walked into it with, but ultimately you found a way to turn that into value for a customer and that’s the sort of attitude that I think leads to whether it’s like a market shortcut or it’s some unique insight or it’s a fresh perspective on things.

[00:09:37] We’ve got three people that are experts and they’re the only ones that can come up with it. Data is exciting to me ’cause, and especially with a data strategy, I want it to be democratized as much as possible. Put it in the hands of people who have a wide set of experiences thought processes and perspectives.

[00:09:56] And then let the data prove out whether it’s a good idea or whether it’s, stupid.

[00:10:01] Jeff Williams: Yeah, that last little part is really interesting. I was thinking before we jumped on that, maybe for the first time, I had taken a moment to think and I was like, I think what my experience working with you thus far at this company has been on data, is that I think the spirit you’ve brought has been a little bit more, and I was thinking about this relative to my perspective and I think the spirit you’ve brought is like this. Well, a little bit less about like the specifics of the contextual data that our customers have and what are the ponies in there, because that’s the stuff that is still fairly new to you and certainly was very new to you.

[00:10:44] But the spirit I feel like you brought was like we can do that. It was almost this let’s do that and figure out what’s going on there because we can, it’s not that hard. And that’s like really refreshing. Especially, that’s why I say getting back to that kind of startup feel that I experienced, it was like yeah, let’s figure that out.

[00:11:04] And then you go through this period of I don’t know. We may not have the resources to do that. And then maybe you have certain people in certain positions that don’t have the same experience and skills and stuff like that. And you whatever the reason is and no judgment against any of those things.

[00:11:20] But the mentality you brought in terms of data is very much yeah, we can do that. And that’s maybe what some of the anxiety, I guess I implied I feel sometimes is let’s you know, do that. And you’re like, Hey, cool, let’s do it. Some of the conversations we’ve had, it’s yeah, we can do that. Let’s do it.

[00:11:39] And it’s extremely freeing and the alternative or the opposite is extremely limiting in terms of opportunities and what you know, what people can do, what our customers can do, working with us, what we can do as a company. All those things. When you take the limit off and open the opportunity, it’s really exciting. Is that fair? Do you feel like you bring that spirit? I want you to know that it feels to me like you do, even if you don’t feel like you do.

[00:12:07] Jef Rice: I like certainly to say yes and figure it out, and then with my approach, I think everybody’s approach should be yes until we find out we can’t I’d much rather die on the hill than decide not to climb at all. And that’s just the approach that I think, what do we have to lose? It’s just data.

[00:12:28] Jeff Williams: Nobody’s dying yet. Nobody’s bleeding yet.

[00:12:33] Very few people have their feelings hurt too. So it’s, why? There’s really nothing to lose.

[00:12:40] And I was gonna say, like to compare and contrast my perspective, I have told you this, I think, but if you don’t hear, here’s this formal in public acknowledgment of it, but I have been, over the last nine months, or 12 months? Goodness, it’s December. However long you’ve been here.

[00:12:58] I have felt certainly re-energized and very stimulated. One, thing that I know very well, and many people that are forced to work with me as well, and the case of my wife who chose to, have association with me though all these people know is I’m a learner.

[00:13:16] I’ve taken these personality assessments, these strengths finders, and stuff like that. And one constant in all of them is I love to learn and I will go down these rabbit holes and stay up all night for days on end. Just because the sort of energy I get is more from the learning.

[00:13:34] And so as it relates to, your time here, the stuff that I’ve learned has been amazing, extremely stimulating, but, that’s not necessarily the thing I like about data. I’m learning about a lot of tools and a lot of processes and stuff like that as we go. But the thing that, you know, to get back to what I feel you bring in the spirit of data and data strategy versus maze. I like finding the ponies in there and coming up with things. And then the learning is this, gravy on top.

[00:14:09] Jef Rice: Yeah, no, it works out. ’cause I feel like the personality tests I’ve taken, I’m more of an enabler. I get pleasure from seeing other people get in there and find the ponies find the treasurer and learn something new. And the more that we can make that happen, the better off we’re gonna be.

[00:14:28] Jeff Williams: Love that. Let’s talk about data strategy generally. You know what, let’s pause that for a second. I want to go back to one thing I was thinking as you were talking about ZipRecruiter is, I’ve heard a lot about this. It feels overwhelming to me that this focus on data and the strategy there was a total differentiator. What was it like for ZipRecruiter customers versus alternatives, what was it that caused zip ZipRecruiter really to separate itself? Was it like the result of the data and the strategy, you were a part of enabling made it better?

[00:15:09] Jef Rice: I think so. Part of it was also us being able to do more for less than the founders of ZipRecruiter. As it turns out are still very risk-averse. And so they wanted to make sure that the things that we were building and delivering to customers had maximum impact. So we didn’t just build on a hunch, we wanted to prove that those things would work.

[00:15:30] So almost everything started from the standpoint of a scientific experiment where you’re thinking like, I’ve got a thesis or a theory, and now, and here’s how I’m gonna prove it, and here’s how I’m going to measure that to show that I’m right or wrong and what’s the sort of minimum amount of work I have to do to get to a place where the data can start to prove that out?

[00:15:51] And it allowed us to rapidly build a set of tools that were different from what you would’ve gotten if you did your typical sort of product management approach to the world, which is let’s start by looking at what all of our other customers have and we’ll put together a big long list and a matrix of it, and then we’ll stack rank it, and then that’s what we’ll go build.

[00:16:11] We didn’t worry about what any of the other competitors had in the market. We just cared about what was gonna move the needle for us with either more visitors that are, job seekers who cared about finding a job or more employers that, cared about finding good candidates. And we were with blinders on and our only sort of guiding light was the data that showed whether the experiments we were running were successful or not. And that made it very unique because we, I don’t think we worried about what anybody else was doing, including, the Indeeds of the world. We were 10 times bigger than we were.

[00:16:48] We just worried about what made our customers happy and what brought them value and made them come back and use our service.

[00:16:56] Jeff Williams: Love it. All right. Let’s talk about data strategy. We talked about what’s exciting about data. You know that this is one of those things. It’s a bit buzzwordy, right? I think if we weren’t having this AI moment, which you referred to, then, data strategy might still be this overused cringe-worthy term being used.

[00:17:15] And anytime you have one of those, if you’re me, you’re like, all right can we agree upon a common definition of such thing before we start talking about it? So how would you define a data strategy? Just very generically, whatever comes to mind. What is it to you?

[00:17:33] Jef Rice: Just understanding all the data that makes up the various systems that you need to run your business. I don’t think it’s terribly complicated from that standpoint. I think it is boring, and certainly, when you approach it from the standpoint of in our industry of the thing that people have historically thought about data, it’s about quality, security, compliance, governance, and if we wanna put the listeners to sleep.

[00:17:59] We can talk about that for half an hour. And it’s boring. It’s and then you get into sort of normalization or standardization, and that’s equally boring. it’s not about the data. It’s about what you can do with the data. It’s about getting the data easily in front of the people that know how to manipulate it.

[00:18:17] I don’t necessarily know how to do that, but you know how to do that, so how can I get our data in front of you, and then you can go and experiment and do exciting stuff with it? That’s the data strategy to me, is how do I get, I’ve been gathering data. In a bunch of different silos, whether it’s CRM or it’s my accounting system, or it’s my portal or my website or portfolio management, or whatever.

[00:18:42] I’ve got all these different silos and all of those things are pumping out data. How do I put it all together in a way that allows me to get more value than I’m currently getting out of it? ’cause I know it works for tactical stuff, but how do I turn it into a strategic asset? And that to me is the strategy piece of it.

[00:19:00] Jeff Williams: Yes. Love that. Let’s talk about some of the things that we’re doing, strategically to enable the possibility of arriving at understanding things that you wouldn’t otherwise from that data when it’s siloed and being capable of doing things and what have you. I suspected there’d be a little bit of this, but didn’t appreciate how much of the sort of buzzword stuff we’d be using here.

[00:19:29] I should’ve anticipated more, but I think, right off the bat, a term that comes to mind that’s been used a lot is data lake. Maybe again, we ought to give a definition. 

[00:19:44] Jef Rice: Yeah. It’s just, in all honesty, it’s not a whole lot different than a database or a data warehouse. It’s just cheaper and easier, and that’s really what spawned the evolution of moving from a database to a data warehouse. And from a data warehouse to a data lake was just cost and a lot of places, and, ZipRecruiter was no different.

[00:20:05] We generated something in the range of a terabyte of information every day. And putting that in a database and having that database server have to be scaled to its physical limits, costs a lot of money. And then where can we put this data that will still allow it to be actionable and queryable and combinable into interesting places, but doesn’t cost a lot of money.

[00:20:28] And that’s where this concept of data lake came in and it was evolved to, I won’t say invented, but certainly evolved to six or eight years ago, and as a way for these big companies that were starting to just pump out data to scale and be able to do that without being buried under infrastructure costs.

[00:20:45] But ultimately it’s the next evolution of a data warehouse where you can bring all your operational data in one place, and make sure that you’ve got high quality. Start treating data like a product instead of just a byproduct of the things that you’re doing. And but while again, like boringly managing security and compliance and governance and all of those things and then get in there and discover cool stuff that you can do with the data.

[00:21:16] Jeff Williams: I want to I think you’ve heard me offer a definition that I’m about to give you, but I want you to tell the listeners too, whether it’s fair and appropriate. I have this visual for some reason with data warehouses, or let you know, that might be unfair.

[00:21:31] Let’s categorize it as pre-data like conceptually this old school idea, right? Like the status quo of what kind of data management may have been versus what it can be now. And some association I suppose to data lakes for reasons I’ll mention here in a second. But I have this visual of data warehouses.

[00:21:52] Are these like I  imagine like Amazon like I see like front loaders and forklifts and everything’s nice and tidy, like such that the robots just zip back and forth and grab the thing. And that’s all cool ’cause that’s like very futuristic and automated and all that sort of stuff.

[00:22:07] But the sort of the what’s cool about the automation, the robots is like you see the cool part there when you’re there, you don’t see all the pain and the heavy lift associated with keeping that place tidy and like staying out of the robot lanes and getting run over by robots and stuff.

[00:22:23] And so if that’s like the visual of the like sort of data warehouse kind of status quo had been my visual of data lakes. Is like a free-flowing thing where it’s like you just dump stuff in there. You don’t have to worry too much and put all this effort into neatly managing all of it and organizing it as it comes in, therefore limiting the potential.

[00:22:49] It’s just getting it in there and when you need something from there, you go figure it out and you know it’s gonna be in there. ’cause it was in there. It just, you didn’t have to spend all the time and effort to define exactly where it was gonna go and how it was gonna be structured and then find out later on that, that limited your ability to do this other thing, with it.

[00:23:09] And then the extending the late concept, like it can free flow out and maybe do other things other places too, instead of being for people to go in there and get it.

[00:23:21] Jef Rice: Absolutely fair. It’s a good, it’s actually a very good visual. The data warehouse is very structured and honestly, the single purpose of a data warehouse is to combine the data in a place that you can get to it easily. And, which is fine, it’s very good at that, but it does take a fair amount of coordination, as you said like I have to have robots and forklifts and all of this sort of understanding of the data ahead of time.

[00:23:48] This does make it more efficient in a database, but it still means that it’s limited in a lot of ways. And I think over the course of the last sort of 10 years, we started to see data as other stuff. It’s not just rows in a database or cells in a spreadsheet. It also is like video. It’s also freeform text, it’s unstructured data.

[00:24:08] It’s all these other things. And so the data lake opens that up to say, bring all of it. It’s way more inclusive. It’s saying everything is data at the end of the day. But we also acknowledge that the nice part about a data lake is that it also builds that process of discovery into the infrastructure itself with a warehouse.

[00:24:30] You have to know what’s coming in because you’re gonna create a structure for it. It’s the equivalent of having to put up a shelf, and then when I’m putting up the shelf, the first thing I’m gonna ask you is what am I storing on the shelf? How big are the boxes? What are the dimensions?

[00:24:44] And then I plan it nicely and as you said, it will end up on there and it’ll be ordered and beautiful with the data lake. It just says to bring the data as it is and the first thing I wanna do is just get it there. The second thing I’m gonna do is clean it up, and then I’m gonna go through this pause period where I just discover and explore the data. I’m just gonna see what’s in there. I’ve never seen it in this perspective before. And once I figure that out, now I’m gonna have a single purpose. I want to use that data for it. It might be an aggregation, it might be a calculation, but it’s gonna end up in a separate table.

[00:25:19] That’s a single purpose, and it will allow the data to continue to grow and evolve in its sort of raw format, but it also allows me to start to build purpose and value on top of that and understand that this is like an ongoing thing. 

[00:25:36] The data lake is much more. Not sure what we’re gonna get, but we know there’s gonna be a point in time where we wish we would’ve collected this data a year ago instead of today. So let’s just get it in there and then we’ll figure out what to do with it and build value in a way that’s isolated and high quality and all of those sorts of things.

[00:25:53] Jeff Williams: Yes. Love it. I’m gonna keep going with this. Metaphor, like you can’t just dump your data off at the door of the warehouse. ’cause the robots will get all tripped up and the whole warehouse will break. And then you find out that the space, you do have to put it in, you gotta get all this machinery and put it up there and stuff.

[00:26:13] And then, oh, by the way, there isn’t enough height for that part of the data. And the business user’s okay, fine we have this timeline we gotta do this on, and we’re already behind it. So just don’t bring that into the Data lake and just dump it, and we’ll figure it out.

[00:26:31] You are only aware of the things you’re aware of. And that’s very limiting conceptually in theory. Versus this just, dump it there and we can know what we know we want at this point, to your point, let’s start to build some flow into it, some direction that the lake is, flowing in. But it doesn’t mean that we are no longer able to, still discover things we don’t know, we don’t know yet that we want to do, and that we no longer can in the warehouse ’cause the height of where we’re storing it, forced us to make a decision to not bring that data.

[00:27:14] Jef Rice: Yep. And that’s, another sort of good analogy is to imagine that you were tasked to cook dinner for yourself and all your friends but you weren’t allowed to go inside a grocery store. You’re only allowed to walk up to a window and tell them, alright, here are the ingredients I need.

[00:27:31] You’re just gonna ask them for stuff you know how to cook, but what if you were allowed to go in the store and just wander around and see what they have and yeah. I’d love to be able to do something with that spice or do something with these ingredients and just play around with it.

[00:27:46] The results from both of those things are like, yes, there will be value, but it’s a very different experience. And with the data lake, it assumes that you don’t know everything you want to know before you enter into this process. And the data warehouse demands that you know, quite a bit before you enter into the process.

[00:28:04] You have to set up the structure and so it, as we talked about at the beginning, it means that experimentation is hard with a data warehouse because the startup cost is really high. Whereas with a data lake, it’s not high. They bring your garbage, throw it all in there, we’ll make some sense out of it and see if we can find some value.

[00:28:26] And if we don’t, no big deal. Maybe we will next week, but we didn’t today. We didn’t lose a whole lot in trying different things out. And that’s what I think is the value of why we spend so much time thinking about data and data lakes and all of that.

[00:28:40] Jeff Williams: And so I think one other thing that we’d be remiss to not discuss here in terms of our ongoings with such things and our initiatives is just the concept of messaging, and I’m gonna continue to extend the lake metaphor perhaps inappropriately if I haven’t already have.

[00:28:59] But I’ll do it by saying there’s, the capability. I don’t know that this is necessary, I certainly don’t want to imply this is exclusive to data lakes versus data warehouses, but I’ll come back and tie the knot on why I think data lakes are better at it. But it allows you to almost set up like entry points into the lake and say, hey, this certain water’s coming through here now. And, then take some sort of action only when that type of water comes into the lake. And then on the other end of it, it could be like when that type of water comes in here, we need to let certain other type of water out and build the rules. Almost like a monitoring system that’s keeping track of it.

[00:29:40] Whereas again, that’s stuff that you could do in a data warehouse. But the need to, really define and structure all the stuff as you, load it into the warehouse, create your space, and use the forklift to put it up where it’s going. And then this other stuff there again, falls into this category of stuff you have to be thinking about. It’s time, it’s money, and it’s limiting oftentimes in the context of a data warehouse, whereas with a data lake. It’s since you don’t have to spend all that time doing that up front, you can at any point say, oh, we brought this water in here. We didn’t even know what type of water it was, and now we are finding out that it’s here and it’s interesting and some, as it’s coming in, when it is that type of water, here’s what to do with it without, a single forklift.

[00:30:30] Jef Rice: That’s right. And the other key piece of it is, and that makes it less expensive to own over time, is that it assumes that you’re constantly experimenting and looking and discovering new value, and you’re doing it in a way that doesn’t disrupt the existing waterways.

[00:30:52] And you want those to be static or reliable. And a lot of times, because of the way you have to combine data in a database or a data warehouse, you only have one copy of that data. But what if I need to slightly modify it to find value in another sort of waterway?

[00:31:11] The infrastructure that exists inside of a data lake protects the raw data. It protects the cleansed data and it protects the aggregate data as separate things. I can have like in number of copies of that aggregated data that provide value, and I can add to those and modify them independently and in a modular way without disrupting all the other waterways.

[00:31:34] That’s what you need to be able to scale is like not only do I need to discover things quickly, but as I accumulate those discoveries and those value points, I need to be able to maintain them in a way that’s not gonna bury me under the weight of that value because I need to, accelerate that innovation in a year just as much as I need to do it today.

[00:31:55] And I think, the best practices that have evolved around data lakes just take a lot of that the last 10 years of learning, of how do you really turn data into a product and into a process and into a science and puts it all into one nice big package for us.

[00:32:11] Jeff Williams: So we’re doing a lot of this, and you and I are working a lot of times on these things. And in terms of what it means to people who might be listening if you are a customer of ours or a potential customer of ours effectively, I guess the simplest way to say it is that we are, our data strategy involves building this stuff out for our own internal infrastructure, within our products, within our operations, all those sorts of things, but extending that effectively to customers to leverage as well. It’s not just a single lake, right? But everything’s connected.

[00:32:45] All the water in the world’s connected.

[00:32:48] So, it does effectively offer them beyond the sort of benefits of the ability for our products to integrate within themselves the ability for our products to integrate with, the rest of the world. Beyond that, it is offering them this, benefit of all those same things we’ve been talking about with really no need for them to do much at all, right?

[00:33:10] Jef Rice: That’s right. And you’re in these conversations as much as I am with our customers, who are searching for ways to do that and whether it’s because they’re starting to scale by having more funds. Or more entities or whether they’re an umbrella company.

[00:33:31] They’ve got six companies underneath. They’re looking for ways to optimize the way they run their business. And that almost always starts with data, and it almost always starts with the idea that there’s some part of the way down the path. Nobody’s starting from oh, I just do this data thing in, in my database.

[00:33:48] They’ve got some semblance of either a warehouse or a data lake, and it’s being able to be, flexible when it comes to how we get them from where they are to where they need to be as quickly as efficiently as possible.

[00:34:00] Jeff Williams: Yeah. And it’s really interesting, right? Because one of the things about the market we operate in is it’s a service. At the end of the day. It’s a financial service that this market provides to their customers, and service industries or markets are really interesting because they aren’t like the let’s experiment with a button here and see if people click it, or, let’s ship a day to our website and, find out what people are doing.

[00:34:27] Now you can, certainly do those things, but I’ve used this analogy before. I did get a little excited and I remain excited about it, but I got a little obsessed with it some time ago about this idea of the impact that software Cloud and now data has on service industries. Especially highly competitive, highly fragmented ones, and even, both B2B and B2C. It’s all about customer experience. This one that I was obsessed with was Veterinarians, right? Imagine the number of veterinarians can, do you know of a single brand? No. Maybe you can’t think of the name now, but the inside Petco or PetSmart, Banfield, or something like that.

[00:35:11] But otherwise, like they are highly fragmented. It’s every corner and the consumer, the pet owner has a million different choices and none of those consumers with those millions of choices are choosing which vet they go to because of like this sort of algo that is on, was developed on the vet’s website and which like forced them into unknowingly, get in your car, turn it on, drive to this location and without knowing and all of a sudden, no, it was like I’m gonna go to this vet and in these service industries, whether it’s B2B or B2C, like all of course, the impression all along is important. But like the experience you have is so key. And so like I was always like we had a Labrador that we took to a certain vet at one point, and we also then took him the Labrador there for doggy daycare and there was a pool and there was another lab and we would pick him up from Doggy daycare and we’d hear about his friend Cooper, and from the people at the daycare. But then we’d go in to see the vet and I was always just sitting there in the examination room watching the vet like on this sort of old-school, huge computer running on like Windows 95.

[00:36:30] And I was always just I can’t wait for the day when this person comes in with an iPad and mentions Cooper, like they had no way of knowing that information, and so like at the end of the day, that’s technology, process data, and which enables people to create a better customer experience.

[00:36:48] Jef Rice: For sure. Yeah, no, that’s what it all boils down to, right?

[00:36:51] Jeff Williams: That’s all this market, really any service, the highly fragmented competitive service market is all about.

[00:36:58] Creating a customer experience. And in this market, there is a lot of data and that’s why we talk about this, right?

[00:37:03] Sure there is a whole thing about investment decisions and stuff like that’s interesting with data given that limited partners pay general partners to make investment decisions, I think having those decisions be informed by data and, AI and stuff like that, and. It’s great. There’s a lot of potential in that, but also let’s not automate that fully, right? Like we’re still using our brains and putting human context into it.

[00:37:31] Jef Rice: Yep.

[00:37:31] Jeff Williams: So what all of what I’m saying is the opportunity with data for people in this market doesn’t have to be a sort of futuristic science fiction film. It can be about being faster more accurate, saving time and pain, and potential error and risk.

[00:37:49] And about creating a better customer experience, whether it’s doing all those things faster and better and with less risk and less, error to produce the same thing you had or about creating a new thing.

[00:38:03] And that’s always been the vision. We talked a lot about it on the podcast behind Altvia is about enabling GPs to create something for LPs that’s differentiated, and not just on time in the same format, but about creating this oh wow, this is more like what I get as a consumer with these massive financial institutions like Schwab and, what have you that are creating these technology-driven experiences for consumers because there’s a lot of choices and it has to be cool.

[00:38:33] And now we enable GPs to do that by way of taking their data and helping them improve decisions and stuff. But again, doesn’t have to be this like a super futuristic science fiction thing. It can be just about providing a better customer experience because that’s what these markets.

[00:38:52] Jef Rice: For sure. Yep, we talk a lot about that. I think we’re in an industry that’s a little bit behind because they’ve been preached to for 10 years about how hard data is, how expensive data is. After all, they’re pulling in consultants to beef up their hours when it comes to that. But the truth is that data is pretty easy especially when you take a data lake approach to it, it’s pretty easy to get it into one place.

[00:39:18] And the key piece is being able to do interesting things with it once it’s in those and I think we get a lot of questions about AI because that’s the buzzword of the year and Altvia has said AI’s interesting. It is, but it’s also a little bit like hiring a recent MBA grad intern who’s like super smart, but has no clue about your business and how we fast-track that AI to get experience, and in the AI world experience is context. So how does it know how to do things in a general way, but it to be really useful, it has to know when and why, and the when and why comes from the data that you store in your business. And being able to centralize that means you can start to layer on things like ML things like AI without the sort of fundamental data lake-based foundation.

[00:40:06] All those things are like little shiny toys and they can do cool little magic tricks, but they’re not fundamentally useful until they understand your business or understand your nuance or understand the way that you in, interact with your customers, the differentiation that you’ve already built.

[00:40:26] And that’s why we really leaned into the idea that data is the key to do the fun stuff. The other aspect of AI is that it’s just evolving at a pace. It’s unlike anything we’ve ever seen. And if you don’t have a foundation, a solid data lake foundation, it’s like having, the latest toys, but you’re gonna break the bank trying to implement those things. Then find out, oh, that’s not supported anymore, or that’s really evolved . The thing that used to work doesn’t work as well. ’cause they’ve changed the underlying large language model or whatever those things are. So for us it’s, let’s get the data all into one place.

[00:41:01] Let’s start doing some interesting things. If AI plays into that, great. If ML plays into that, great. Sometimes it’s just knowing who to call next. What’s the next thing to do? And that doesn’t necessarily come from AI and sometimes it’s putting the right information in front of the right person. 

[00:41:24] Being able to bring up a video that shows them playing together is something that would differentiate that experience. And I think that all comes from data and context that you’re not gonna get without doing that sort of foundational work.

[00:41:38] And we’re focused on figuring out how to make it cheap and easy to do it so that we can break that sort of industry belief that data is hard and slow and expensive. So that we can get onto this sort of interesting, cool stuff of what do we do with it now that we’ve got it in one place.

[00:41:57] Jeff Williams: We talk about this a lot, man. And then we decided that we would record it and put it on the podcast, and not surprisingly super stimulating just to do it, formally here. There’s all sorts of thoughts and all sorts of analogies and metaphors and all sorts of things that really, get get you excited. 

[00:42:19] I think that’s representative of how we feel on a day-to-day basis and stuff that we’re working on. But as I mentioned, there’s a big part of me that’s stimulated by learning things. When we set out to build this company years ago, it was like the personal thing to me was like, I don’t think I knew that about the learner thing.

[00:42:37] I think it was present to a lesser degree then, but, the point is I still remain, driven by this. But the impact that this can have on this market is again, in some cases pretty basic. It’s enabling you to create a better experience and that’s this thing because, early in my career it felt like, man, it’s pretty amazing how much money is moving around this market and what this service actually entails.

[00:43:04] It’s, a little mismatched in my opinion. But we brought in people over the years and been a bit younger and they feel this desire to be connected to their work. It occurred to us that they aren’t tied to that. And then it was what becomes possible in the world if, the sort of market we’re powering is this very quickly moving, very dynamic, and very influential capital source that is, lesser known to most of the public in terms of being private markets.

[00:43:39] But if we can make the process of LPs and GPs coming together and having a coming together more efficiently, but also having a more efficient and dynamic relationship and then powering the GPs to not have to worry about that so they can focus on other things. And then also the technology can help them focus on better decisions and better processing and stuff. With technology and data, what becomes possible is that big of a stretch to think that maybe it can power the world’s, next greatest innovation, maybe like a cure to an unmet disease, or create the next company that solves a climate problem.

[00:44:23] Whatever it is. It’s I don’t know, and we don’t have to know what it does. It’s reasonable to assume that if we play that part, that is a possibility and great. You know that helps me get fired up.

[00:44:38] Jef Rice: Yeah, for sure. Same here. You’re talking about time. It’s a fixed resource and our customers have a limited amount of it and they can deploy it, dealing with their LPs and strengthening that relationship. Or they can spend it figuring out how to deploy that capital in meaningful ways, and we’re trying to figure out how to get them to focus their time away from the menial aspects of running a business and more on the pieces like you said, that you can’t automate that. You need a person in there who has that context and experience, and it’s really around deploying the capital, figuring out how to change the world through the investments that they’re making.

[00:45:18] Jeff Williams: Brother, this has been a pleasure. Very fun. Thank you for coming to work here, man. Thank you for having the spirit you do for taking the positions you do and saying the things you do. It’s a joy to work with you and have you be a part of the team, but also to sit here for a minute and kinda formalize talking about that. 

[00:45:40] Jef Rice: Yeah, no, it’s been fun. I wasn’t blown smoke when I said, this is a great team. I’m really happy I made the decision and haven’t regretted it for a minute, and looking forward to the things we’re gonna be able to do together in 2024.

[00:45:53] Jeff Williams: Feelings. Very mutual brother. Thank you and we’ll get on with it now.