Latest / CIO Unscripted / Sell, Support, Ship: Startup Lessons and Using AI with Jeremy Brown
Transcript
- Jeremy: people are so enthralled with the technology, they forget that nothing happens unless you can get that first customer signed. And nothing happens until you can get the second one signed by keeping the first one happy. So all the all the cool stuff you can build, and that's one thing I see a lot of startups struggle with, is they're so busy. being impressed with their own technology that they forget I've got to go sell it and I've got to support it and get it out the door. And then the then the last bit, so it's gotta sell it, gotta support it, and then you gotta ship it. And so many people just just a very simple formula. sell, support, and ship. That's what you do all day long. And if you can, if you can execute on that and get it out the door, you have a huge success rate in startups Niel: Welcome everybody to CIO Unscripted today with Jeremy Brown. We thank VLCM for sponsoring this for us. Jeremy is a little bit different from some of the other technology leaders that we have had on the program because I call him sort of a startup guy. He does startups. We haven't had a lot of startups on here. We've had mostly like enterprise CIO CTOs. So I want to talk to Jeremy, about that, about those experiences, how he got started in that, what he's learned, and what are the kinds of because apparently this AI thing is going to be around for a while. I've kind of finally given up on the fact that it's a fad. So from Jeremy: Yeah. Niel: a startup perspective, what are the kind of use cases for AI? And then we'll just kind of see where it goes from there. So with Jeremy, I'd start with Jeremy, I'd love to have you just kind of recap your career. How you got on the path you are and some of your more memorable experiences as a technology leader. Jeremy: Well, thanks for having me today. I'm really excited to be here. I have an odd career. Now, let me say it that way, because I love to work on the stuff that nobody else wants to work on. And so I kind of started off as an engineer and I was working on sysadmin doing stuff for data centers. You know, I got my start at a little company called Barrio, and we were doing data centers and we were helping with servers and doing sysadmin work and troubleshooting and building stuff and then worked my way into QA and I loved my QA experience because I was automating the heck out of things. And so I graduated saying, I want to be a programmer. I want to be a programmer. And I've spent half of my career doing everything else but programming. So it's been an odd, it's been an odd journey because I've done the data centers. I've done QA. I've, you know, managed a lot of the data sides on top of that. And so a lot of people just kind of stayed to software and they stick just with software and they don't do anything else. And so as a result, me kind of hitting all the other areas, it really helped help helped me get to where I was a CTO because I knew all the different areas I needed to work with. And the thing I love to do is I love to get systems that talk together. And that was one of my favorite things to to build in a company. Come in and get it so the financial system and the software were actually talking. And 10, you know, 10, 15 years ago, that was, you know, light. I mean, that was amazing. Niel: That was work. Jeremy: Rocket science. It was hard. It was hard stuff. And Niel: That was work. Mm-hmm. Jeremy: over the years, it just keeps getting simpler and simpler. But the thing that happened is in that time, we now have data is so critical to every part of the organization. Where all of a sudden I'm I'm working on data warehouses, whether I want to or not, because everybody needs access to the data. And so it's kind of funny. I started with hardware, worked into software, did QA for quite a bit, then went back to software. Then worked my way into the cloud and I was doing Amazon and Google and Azure and and just kind of doing those. And it's funny because I I did the big companies and I did the huge companies and I kept going back to the startups because I just love being scrappy and being able to work on everything and start from scratch and kind of working my way up the up the up the chain. And it's funny because I I did quite a few startups along the way, you know, some were great success, other ones not so great, but had a lot of fun along the way. And then finally the AI, I just stumbled into it because I needed it for data. I I kept running into it, you know, five, six years ago. I was kept running into I need to use AI for data. I need to fix this thing with AI. I need to fix that with AI. And I I was hesitant at first. And then I just realized it makes my data life easier. I should definitely use this. And then immediately it started turning into, I can use AI to solve all these other problems. And it's funny because I keep getting the questions, well, is it going to replace your job? Are you going to lose your job? And I said, no, everywhere I go, it creates more jobs. And I'm able to build more software and do more integrations and do more of the solutions. The one thing I do like it does is it helps me get rid of the people doing busy work. The data entry, the people manually walking papers across the office or collecting papers or doing, you know, just the stuff that people were doing just to fill time. Because we need somebody to take care of that because we can't do it ourselves. And I was talking to someone just the other day where like, my gosh, I can do this financial analysis in hours where it used to take me weeks. And I'm like, yeah, that's that's what I love about it. And it helps me to go faster, but I I'm still finding a lot of people are struggling with the critical thinking skills. They can't, they can't put their arm around what do I want to build? And I'm like, no, that's what AI it AI needs you to give context and give all the details so that you can get an output. For me, it's been kind of a it's a great journey. I love the data side. I love the the cloud and data center side. I love the software side. And so I kind of bounce between them all the time, just kind of going back and forth. And with startups, you really have to be, you know, wear all the hats. You have to do everything. And so that's the part I enjoy the most. It's just kind of sitting there going, I could I could do that. Hold on a second. Let me figure that out for you real quick and get it done. So that's kind of my background and what I love and how I love to do it. And I've had great, I mean, great teachers all along the way. People that just kind of, no, you can't do it that way. And I'm like, thank goodness you showed me that one. I can't do that. You're right. And just different people that had the experience to kind of push me in the right direction. No, do this instead of that. Do this instead of that. And so it's been a lot of fun. A lot of architecture, a lot of Niel: Yeah. Jeremy: consulting along the way and getting things done and solving problems. Niel: So what do you think as you look back over your career? I'm talking about at the startup level, at the organizational, at the business level. What do you think are the critical success factors? You know, I've got the I've got this amazing idea, I've had it for 30 years, and I'm ready to to launch my startup. what would you tell me to make sure I do? Jeremy: the thing that I'm kind of weird and unique is I am a hundred percent focused on sales all the time. Niel: Okay. Jeremy: Because people are so enthralled with the technology, they forget that nothing happens unless you can get that first customer signed. And nothing happens until you can get the second one signed by keeping the first one happy. So all the all the cool stuff you can build, and that's one thing I see a lot of startups struggle with, is they're so busy. being impressed with their own technology that they forget I've got to go sell it and I've got to support it and get it out the door. And then the then the last bit, so it's gotta sell it, gotta support it, and then you gotta ship it. And so many people just just a very simple formula. sell, support, and ship. That's what you do all day long. And if you can, if you can execute on that and get it out the door, you're you have a huge success rate in startups Niel: I I once my startup experience was kind of coming in to help clean up a challenged startup. And one of the biggest issues was the founder was an idea generator. Jeremy: Yes. Niel: And he had a hard time getting to the ship part because Jeremy: Yep. Niel: he'd already moved on to the next idea. And the way we turned it around was saying, okay, you can have those ideas, and they'll go on the backlog. Jeremy: Yeah. Niel: And we'll we'll implement them after we get the thing we're working on already done. Do you see that much? Jeremy: Yeah. All the time. All the time. People are so excited to chase the next thing. And right now with AI, that's what's happening right now. So right now, I'm a chief AI officer. And I Niel: Okay. Jeremy: we are basically automating all the internal systems so that we don't have to have as many employees, but at the same time providing solutions for customers that they don't have to come to support. They can just take care of themselves. And we build the you know, we build the right pieces. And we're not even doing anything too elegant. We're just trying to put the pieces together to help them deliver and have better value. And we're focused on fintech. So we're doing a lot of payments. We're doing a lot of pieces, but we're making it so it's a better experience. But that happens all the time. The founders are just so excited and AI, they're so excited about the next tool and the next tool. Niel: Sure. Jeremy: And they they don't even want to finish what they've started because they're like, no, I got it, I got it. I gotta be in I gotta be in front of that. We gotta do that point. I can Niel: I and I can make it better by doing this. Jeremy: I can make it better. I can make a change. We can write more code. And I'm like, nobody needs more code. You need reliable, shippable code. That's what everybody needs. But everyone's so excited about building so many new things that they forget about, but you gotta be able to maintain it afterwards. yeah, that thing. But that's the part that's it's really important to be that bulldog that just kind of latches on and says, I'm getting this thing across the finish line. I gotta deliver it. I gotta get it out the door. And that's the part that startups they they struggle with that exact piece because they're they have lots of money. They want to chase the whatever dream they can think of, and they're gonna go as fast as they can. And in the end, it's just like, it's a recipe for disaster because you're never gonna catch it. You're always gonna be chasing chasing the next thing if you do it that way. Niel: So that brings up another question. What lessons from a startup can an enterprise technology leader like me? I've only done, you know, that startup experience was 30 years ago. I hope never to repeat it. Jeremy: Right. Niel: It just wasn't for me. Jeremy: the thing that I love the most about my enterprise experience is just the consistency and the process that you can bring with you. And people always think that, I I can never do a startup. I've been in an enterprise my whole life. But you come with this knowledge of how do all these things are supposed to work? And you can bring people along the way to get there eventually. And that's the part that I also see a lot of startups. They all come in fresh and nobody's ever seen it happen. Nobody's ever seen all the pieces put together. And they have a brand new team, and then they're like, yeah, we're gonna make this happen, but nobody's ever been there. And the enter the guys Niel: I see. Interesting. Jeremy: coming from enterprise has said, no, I know it's gonna be. We just have to do these 10 things to get there. And then all of a sudden, it's like, I see what you're saying. And it it's it's funny because if you've never been there and you're all brand new green, you've got to stumble a lot. But if you've been there, you're like, No, no, this is you don't have to do all that extra work, you can just do these things. And that'll help you get your projects in the right state for project management and for the backlog. it's funny because one startup I it made me laugh so hard is I walked in and I came in a little bit later. They had been there about a year, and the insanity they were doing inside their backlog and how they were managing their backlog and their scrum process. And I just said, stop. And we set up a brand new project with the simplest possible statuses. With the very simplest pieces to use. And we said, okay, use that. And most of their problems went away because they were trying to be so elegant and Niel: Yeah. Jeremy: send all these messages and get these approvals and do all these different things. And I said, That's if you are a 10,000-person company, you need that in place. You need Niel: Interesting. Jeremy: these little things. But it's funny because you know the steps to get there. When you set up in an enterprise, when you set up a brand new team, you do the exact same process over and over and over again. And that's what you're doing at the startup. So People coming from Enterprise have a huge advantage because they've seen it. They know how it works. They know what a good deployment is. They know what a DevOps team should do. They know what a QA team should do. They know what the sales engineer should do. They know. And so it's funny because companies that have nobody with that experience, man, they are just making it up and just kind of throwing things around, trying to figure it out. And and honestly, it's it's funny. The one thing they seem to mess up the most is just a good customer experience. Niel: Yeah. Jeremy: They're so busy trying to be something for everything instead of just set a process and make everybody reasonably happy. But they're so busy trying to make everybody. Niel: And so I'm so busy trying to be cool with the technology where it just needs to work. Jeremy: Or they're trying to provide this the most excellent customer Niel: Yeah. Jeremy: support experience that you could never scale. You know, it's like you can't do that. You there's no way. You have to provide a good experience that people say, Yeah, they're a good company. I want to stay with them. Niel: So when you talk about that there's a role for an enterprise person in a startup, does that include sometimes helping the founder think enterprise? Okay. Jeremy: Yep. Yes. And that's that that's the part that's there's a special breed of founders that can be entrepreneurs and they want to do it, but they bring along a lot of baggage with them. They really they're trying to do so much with so little and doing all these different things at once. And as they start growing faster, they want to go more and more and more. And so you're you're trying to just kind of help them be disciplined to say, no, this is how we get there. Let's make a plan and let's deliver in a year. And it's funny because none of them want to make a plan. That's that's something I always all the founders for startups, they're all just like, I can't plan next next week. And I said, They're Niel: They're skilled winging it people, right? Jeremy: winging it people. That's exactly right. And so you have to say, No, we need to set a plan. And just even a three month plan will make light years of difference compared to, we don't know what we're doing next week. Niel: All right, so Jeremy: And Niel: that's where your life and mine intersect or overlap Jeremy: Yes. Niel: because I've worked with plenty of CEOs and executive teams for established companies, wildly Jeremy: Yes. Niel: successful companies, that are still winging it every day. Jeremy: Yeah. No, that's true. That's it's very true. And you have to spend time. Let's be disciplined. Let's deliver the value. Let's fit stay on target. Let's get the it's true. Yeah, everywhere I go, I have to fight that battle. But founders are just, yeah, they're extreme. Extreme winging it. That they're just trying to make payroll. And then once they start to get a little bit of success, they can't turn it off. And so you have to kind of, no, we have to now go. We've now done the winging it part. Now we need to go to the Little bit of process and a little bit more and a little bit inst kind of work your way down that. But yeah, I I have to say there are a lot of CEOs and and leaders that make me do it. They're just like, we don't have time for that. We've got to just change everything. And you're like, Okay, here's the price tag. And immediately it's like, we can't afford that. And I'm like, Okay, then let's make a plan. Niel: So in the world of startups, how do you build a team? You're fairly resource constrained, I'm guessing. Jeremy: Yes. And so you have to find the right people. And I love, I love the old adage of get the right people on the bus Niel: Yeah. Jeremy: because it's very important. You cannot have one-trick ponies. That's the most critical thing. You cannot have a guy that only does UI. You've got to have a guy that can do the DevOps and the UI and work in the database. And so for startups, I'm always trying to find the people that are more generous and they can work across lots of different things. When I go to enterprise companies and bigger companies, I can have the specialists everywhere. But when I'm when I'm in a startup, I'm finding people that are they're willing to do everything and then they have the skill set they can do anything. And you're just kind of saying, and and I I have to spend a lot of time breaking teams down and saying, we're all generalists and we've got to get across the finish line. I don't care if you're a UI guy, you've got to be a back-end guy and a database guy too. And then getting those people so they're all just on the same page. And if they aren't, Get rid of them. Get rid of them fast and just find somebody that will do it. Because Niel: Interstanding. Jeremy: the having the wrong personality in a startup, it's it's death. I mean, you will not get soup Niel: Yeah, I would imagine it's pretty toxic if you got the wrong people. Jeremy: super toxic and you have zero margin for error. You have to deliver and get it Niel: Yeah. Yeah. Jeremy: done. So that's the piece that I always liked is I say getting the right people there that want to do it. And the funny thing is, is most people are super talented. They can find another job. It's not like you're you're destroying their lives. You're like, hey, this isn't work. And the best part is I don't even have to, I just say, hey, this isn't working. Let's help you find another job while we replace you and we've got to get you in the right spot. And and it's amazing how many times they're, okay. Yeah, I get it. Yeah, I'm not really gelling well at the team. I I feel it. It's fine. And they usually go on to bigger and better things anyways. And it's there's some personalities that only work well at big companies. And there's other personalities that they love startups. They love that stuff. Niel: Yeah. I agree. Jeremy: Figuring out all the all the different pieces and wearing lots of hats. They love that stuff. And they're they're willing to jump on with customers. customers haven't problem. Yeah, then you're let me go let me help No problem. And there's a lot of guys that won't do that. Niel: Yeah, there are. Especially in technology. Jeremy: especially in technology. But Niel: Yeah, I've had plenty of people on my team that I would say, I would never put this person in front of a customer. Jeremy: my gosh, yes, yes, I've had them. But they're super talented, like so Niel: Yeah. Jeremy: good and so amazing. But I'm like, yeah, you can't present that. Niel: One of the best cyber people I ever had on my team was I'm like, I put him, I had him one the customer call once going through some compliance checklists. And afterwards, I said, I'm sorry, I can never put you in front of a customer again. Because it was like, Do you do this? No, we don't do that. I said, wait, wait, wait, wait, wait. We do do this, we approach it a different way. We approach, we achieve the outcome, we'd use a different method to get to the outcome than the one you're asking about. Jeremy: Because they're asking in a business sense, not in a actual precise technology question. And that and Niel: Yes. Yes, exactly. Jeremy: they can't make the dis well, it's funny because those guys I usually say, I can't have you do that anymore. They're like, Good, I didn't want to do it anyways. Niel: Yeah, no, yeah, it was exactly what happened in this case. This guy hated it hated talking to anybody except another cyber person. Jeremy: But they're so good. And you have to have them on those big teams. I need to have that specialist that can figure out how to do the DevOps to the I mean the most precise way possible. But on the small teams, I'm like, no, no, I just need a quick little get it done, get it shipped kind of approach. And so that's no, I was gonna Niel: So what does go ahead, sorry, cut you off. Go ahead. Jeremy: say, but that's the thing, finding those right personalities is but is always about. Niel: Yeah. as you know, I'm not that smart. And Jeremy: Whatever. Niel: I've heard this team this term AI native companies. What does an AI native company look like? Jeremy: I like to say when an AI native company comes along, they're aggressively automating their SLs out of jobs. It's not, it's not that any of the technology is any different. They've just set up enough pieces that they're aggressively saying, How can I get rid of this function permanently? Niel: Okay. Jeremy: So, like s so for us, the the one that we're doing right now is we're aggressively getting rid of. The processes that sync data between the financial system and the product. We're just like, we're trying to get rid of it completely so it's transparent. It just works on both systems and everything is just done automatically. We never have to touch it, we never have to think about it. The dashboard's done, the syncing's done, the system just works natively. And we know the pieces are, we know how the pieces work on both sides so well that we never have to think about it. Niel: Okay. Jeremy: And that's the the AI native I always like to think of. They've just taken the time to build the infrastructure so it is possible to do that. Niel: Okay. Jeremy: And so many times people are like, well, we're an AI company. And I said, you put up one thing to solve one problem. I've put up 10 systems and it solves hundreds of problems. Because we're Niel: Mm, okay. Jeremy: really trying to say what can we get by with and what what do we need people for? And what can we just automate that so we don't ever have to worry about it? And so for a lot of my financials, I'm just able to say, we have a software that takes that and it's it has a lot of AI built into it. And then I have a lot of AI connecting it. And everyone always talks about how incredible AI is. It's not the AI that's incredible. It's the pipes to make AI incredible that are it it's building the pipes and connecting the pipes. That's that's 90% of the job. And they always say, Well, I I just did this really cool thing in Claude. And I said, You only did that in Claude because I spent six months connecting every single one of these systems and set up a dashboard so that you could see what you could see, and then you knew what questions to ask. And so, and Niel: Yeah. Yeah. Jeremy: the funny thing is, is they're they're talking like, AI made all this happen. I said, No, we've been doing it for 10 years with data, making sure everybody could see all the data is in the data lake house. And if the data is in the lake house, then boom, I can do anything I want. And that's what AI is living off of that dream of data transfer. I mean data flowing between all the Niel: Yeah. Jeremy: different systems. Niel: Yeah, the data engineering aspects of it that before were done primarily by brute force. Jeremy: Yes. And now and then they talk about the MCPs being this is this so much more I said, it's a it's a fancy API that we were already setting up and getting working. We just put a different in you know, getting it all connected. And I said, Yes, it's incredible what we can do, but it's only because we've put thirty years into it to the point where you can actually do that. And that's the part that I Niel: Well, you know, you know, I've I I'm pretty sure every vendor on the face of the earth now has an MCP server, right? Offers an MCP server. Jeremy: Yes, yup Yes. Niel: And you're right. It what is its functionality? It's an API, it's a connector. Jeremy: It's a it's an API and all the AI is doing is looking for normal patterns and displaying the right report. And people Niel: Yep. Yep. Jeremy: I I they hate they hate when I say that. I said, all AI is doing is it's just it's displaying a report that someone built five years ago and taking advantage saying, look at this incredible thing that AI built. I said, Yeah, but a developer had to figure that out five years ago, and then that's what you're using now, and it's just reproducing the exact same thing over and over. Niel: To what extent are you using AI coding platforms for software development? Okay. Jeremy: quite a bit. we have all of our teams enabled. We have them using it. I I always like to tell people, I said, I don't like a hundred percent API, our AI driven code. I said, I love taking frameworks, building AI on top of it for the pieces I need, and then delivering a solution. I said, because it's this it's the same problem I had. People that would write code from scratch and people that would use library. Niel: Yeah. Jeremy: And it's the same thing. I said, use the frameworks. Build on top of it, have a great solution, and I can maintain that. But if you kind of write everything from scratch and say, I built my own CRM from scratch, I'm gonna look at you and say, Wow, that's not very smart because there's about 30 platforms that have 80% of it done, and all you had to do is just build on top of the APIs and be done and have a solid plot product. there's a couple new products coming out in the last year that are just they're they're all trying to be headless. They're like, I don't even care about what you're gonna build. I'm just giving you the framework and then you can kind of just pick and choose how you want to lay out on top. And I said, Yeah, it's brilliant, but it's based upon all the all the work that we put into it over the years. And Niel: Yeah. Yeah. Jeremy: I remember the first machine learning project I did, and this was years ago. This really dates me. And it was incredibly hard. And I had to pay so much for hardware, I had to pay for the software, I had to pay for the special engineers, I had to do Niel: Yes. Yep. Yep. Jeremy: all this different work. And I remember then the second time I did it, I was just like, whoa. This is so much easier. And then the third time it was so much easier. And now it's just like it's just native. People have it and it works. And it's just so much easier. But I remember having to get there what we had to do. And it's the same way. How I installed the PC the first time to how I install a PC today, Niel: yeah. Jeremy: it's night and day. It's not even the same experience. Niel: Yeah. Jeremy: So it it's lots of fun. I do love it and it's so much fun. But the AI coding, I mean, we use the skills all the time. And I'm using it in report generation. I'm using it in writing code. I'm using it in writing these pieces. it finds all the stuff I need in minutes where it used to take me 30 minutes to find it. Now I find it just a few minutes and I can get it all right, do it the way I want to do it and build the stuff I want to. And so I spend a lot of time going fast on the front end and I spend a lot of time making sure the back ends are solid and obscure and in the way I want them. And so that's the part that's a lot of fun. I love I love the AI tools. Niel: besides the focus on the back end, what other what are some of the other ripple effects of using AI coding in your software development lifecycle? Jeremy: I use a a lot for security. It's Niel: Okay. Jeremy: it's such a great check for saying what are yes, looking for Niel: You mean just looking for t vulnerabilities? Okay. Jeremy: things that are common things that they have. It's just one big searchable database that knows exactly where to look. And Niel: Yeah. Okay. Jeremy: the part that's so funny is I've heard a lot about prompting. I use a lot of the prompting and stuff like that, and how to secure the prompting and stuff. It's just another interface that I have to secure. And so I had to spend time making sure I'm hitting every interface and then bringing back the security. So it helps me with security quite a bit. I don't feel it helps me as much with QA. It does really good smoke tests with QA, but I still have to make sure the logic makes sense. And so I'm still peop good critical thinkers still have to use that part and make sure I'm building the right thing. And so for me, it excellent for interfaces, great for security. Really great for helping me to kind of get the API published, if that makes sense. Niel: Yeah. Jeremy: But on the back end and the and the right functionality and the flows, I still spend my time, most of my time there, making sure that the I built the right API, I built the right flows, it has the right user experience, and it's what they need to be successful. Because you can build UIs all day long, but if you build them wrong, nobody's gonna buy your product. Niel: you know, I've as I've thought about it and I've talked to others about it, there's AI's good at code generation, it's not yet good at architecture. And that's I I think that goes along with what you're saying, where you've got to really think through the design principles and how it all works together. Jeremy: Well, it's it's funny because I spend most of my time asking those hard questions. Well, what happens if I need to scale the million users? Is it gonna work? Niel: Yeah. Jeremy: And immediately they say, No, it's it's not gonna work. And I said, Why? And they tell me the reasons and I said, Well, why didn't we just solve that problem now instead of waiting six months to solve that problem? And they immediately they're like, yeah, you're right. We need to we need to think about throughput and, you know, scaling and how do we get all the pieces. And I said, you just have to think about what is it gonna run on? How am I gonna load balance across it? How am I gonna make sure everybody has a good experience? And how would I guarantee they're gonna have the same experience? And it's not gonna be first guy's a second, third guy's 30 seconds, you know, that kind of thing. And so Niel: Yes, right. Jeremy: that's that's the piece where the architecture and the just laying it out correctly that you can scale and you can do the performance tests and it's secure and you've thought of the things. Because the number of times people have set up systems and I said, You you can't do it that way. No, we're not, we're not putting that on the internet. There's there's no way I'm putting a thousand APIs on the internet. I'm gonna put one API on the internet Niel: Yeah. Yeah. Jeremy: with some features and then the rest of them are gonna be hidden behind by firewall because I don't want that security risk. But it's it's just a it's a mindset and you you do that at enterprise just naturally. And little companies are like, Well, I don't understand. Why can't I just put everything out there? And you're like, No, you wanna make it somewhat secure. Niel: Yeah, and maybe it's goes back to what you first talked about. It's the it's the rush to produce. Jeremy: Yes. They're going so fast they don't stop to think in six months this is going to be a nightmare. I had I had had one company that they actually set it up so every customer was their own cluster. And it was fine when they had 10 customers. Niel: Sure. Jeremy: Then they reached 10,000 customers and they had 10,000 clusters. And I looked at them, I said, So what are we gonna do about it? Because we can't run 10,000 clusters, guys. Niel: No, you gotta re architect the whole thing. Jeremy: Yeah. So they they basically just took it and made it so every cluster became multi-tenant in two big clusters. And we were able to kind of, you know, just re-architect Niel: Yeah. interesting. Okay. Jeremy: and get it in there pretty simply. And we took the same technology, and all we had to do is just put a very simple, this is the tenant ID. This is how they see their data. This is how we segment the Niel: Yeah. Okay. Jeremy: database. But it was easy. It was easy just to set it up as a cluster. And I just laughed. I said, Yeah, it's super secure and really easy. And then you get 10,000, and then what are you gonna do? No, Niel: Yeah. Jeremy: I I my favorite thing was watching one software where they they were talking about how incredible it was, how great it was, and then immediately I looked at the back end and I said, Wait, does everybody require their own login server? And that's that was the technology that it built. And I go, we can't do that, guys, because I can't secure hundreds of login servers for every single customer, but that was the way they solved it. They didn't want to, Niel: Yeah, I'm just Jeremy: they didn't want to figure out the hard enterprise problem. They wanted just to, hey, let me just get this going. This will be great. You're like, no, it won't be. But it's fun. It's it's a lot of fun. And the part that's really interesting is everyone thinks every industry is the same. And that's I so I have another weird thing that I've worked across lots of different industries. I've done identity and security, I've done fintech, I've done education, I've done the sporting events, I've done, you know, ticket sales. I've I've done lots of different companies. And everyone thinks theirs is the hardest industry. And then everyone's pretty similar. And it's exactly opposite where there are a few really hard industries that have very difficult technologies, but most everyone's has nuances. And most of the time they aren't that difficult. You know, it's like, yeah, you're just You're just pretty common. I said, there's some that are just like, my goodness, that's the worst thing I've ever done in my life. And I've I've had a couple of those where I'm like, that was really hard. But it's it's funny how everybody thinks they're very unique. They're they're I'm so different. I it everyone has unique problems, but similar problems if that's if that makes sense. Niel: Yeah, one of my one of my CIO roles, there we weren't financial services, but our largest customer customers were financial services. So we had to, we inherited all their compliance requirements. Jeremy: ohh Niel: And the team was just struggling. I was saying, here's my rule of thumb. If my mother knew we were doing this, would she approve? Meaning, I'm going to protect Jeremy: Like that. Niel: their identities, my mother would approve. I'm going to protect their data, my mother would approve. You know what I mean? Just it it had nothing to do with financial services. If we're doing the right things, compliance should be a foregone conclusion. Independent of industry. Jeremy: Yeah. It's just it's just check marks to say yes. Yeah. Niel: Independent of industry, if I'm doing the right things, compliance should be a foregone conclusion. Jeremy: I have to say one thing I learned along the way, it's easier to secure less data than to secure everything. Just you don't have to hold all the data if you don't need it. Just don't do it. Especially in the Niel: You don't? Yeah. There's a Jeremy: financial services where I'm just like, No, if I don't need it, I'm not gonna store it. I'm gonna let you store it and I'm only gonna store just what I need and not it. No, that's Niel: Yeah, exactly. Yeah. Yeah, there's there's this wonderful practice called triaging things and deciding what deserves what instead of er everything Jeremy: Yep. Niel: deserves everything. Jeremy: no, I the number of times people come to me, I'm like, I need to encrypt the whole database. And I'm like, the whole database? yeah, we have to encrypt the whole database. What about logging? no, the whole database. I said, How about we just don't put sensitive data in the logs? Wouldn't that be better? And Niel: There you go. Yeah. Jeremy: they're just like, no, no, we have to encrypt the whole database. And I said, Okay, here's the cost. And immediately it changes. Yeah. my, Niel: Yeah, especially for the logs. Yeah. Jeremy: especially for logs. But the number of times people were just lazy and they start throwing sensitive data out to the logs, you're like, Whoa, is that a credit card? no, Niel: Yeah, yeah, yeah. Jeremy: no, no, no, we don't ever put that in the logs. Niel: Yeah. Anyway, yeah, it's well, you've had an interesting career. and I I'm not jealous of you at all 'cause it sounds sounds like way more chaotic than I would ever enjoy. Jeremy: It it is, but it was fun. We had a lot of good times and I learned a lot of things what not to do along the way. And so it made a lot more fun as I do it. The times it is successful. I'm like, I get learned from all these other things and it's great. Niel: Right. Yeah. Which I think is generally true of our careers. We Jeremy: Yes. Niel: we we make a mistake and then vow to never make that mistake again. And if we're successful, then we're successful at not making that mistake again. And over time we get better and better because we're making Jeremy: yeah. Niel: not making those same mistakes over and over. Jeremy: There's a couple people that always say it's it's not the experience of what to build, it's the experience of what not to build. I'm like, Niel: Exactly. The Jeremy: I know all of those tricks. Niel: the harshest lessons of my life were failures, and that's where I learned the most. I said, Jeremy: Yes. Niel: I called my Scarlett O’Hara moment as God is Jeremy: Yes. Niel: my witness, I will never go hungry again, I will never let this happen again. Jeremy: Yes. That's exactly right. And it's funny because I get a lot of trouble because I stand up and I say, We are not doing this, guys. Niel: Yeah. Jeremy: I've been here. I know how this ship sinks. We are not doing this. And people get Niel: Yeah. Yeah. Jeremy: really mad at me. And I said, But it helps them to be aware. And they're just like, Niel: Yeah. Jeremy: No, no, you're right. We don't want to. I'm like, I've been here. I know how this is gonna end. Let's not do this. Let's find another way. And honestly, instantly people, if they stop thinking that, they go, yeah, there are a couple other options. And I'm like, Yes. Niel: Yeah. Yeah. Jeremy: Let's try one of those and they work out they usually work out very well. Niel: Thanks, Jeremy. It's been Jeremy: No problem. Niel: great to talk with you. So we've spent some time with Jeremy Brown, a startup guy, and learn from him on some, I think, some important lessons for any technology leader. Lessons from a startup guy on how to scale, do things right, leverage AI in an appropriate way, and make sure we're focused on delivering. Jeremy: Yes. Niel: The right thing as soon as possible. But the Jeremy: As soon as possible. That's exactly Niel: right thing, rather than the thing we don't have yet that we'd like to have, and the thing we've already got done we don't want to release yet, because I know I can make it better if you just give me three more months. Yes. Yeah. Jeremy: I lo I love that it has to be just good enough. And that deliver that and people will be happy and they'll pay you money. That's what you're looking for. Niel: Well thanks Jeremy. Any other final advice? Jeremy: I think that's it. Thanks for having me today. It was lot of fun. Niel: All right, well that's a wrap for today's CIO unscripted with Jeremy Brown. Jeremy: Thanks, Niel