Can a lean engineering team of three modernize the charter bus industry? Brady Perry, Co-founder and CTO of Busie, shares how they are replacing fax machines and paper checks with a modern operating system for group mobility. Discover how real-time data streaming became the backbone of their platform—and how an event-driven architecture with Apache Kafka® enabled a small team to decouple services, stay resilient, and ship value faster with Confluent Cloud.
How is it possible to order groceries to your doorstep in seconds, yet still have to fax a signed contract and deliver a physical check to book a bus?
For Busie, the answer wasn’t just building a marketplace, it was building the underlying infrastructure from the ground up. By adopting a central nervous system approach to data, their lean engineering team has modernized a legacy industry, replacing manual workflows with real-time ecommerce and automated operations.
In this episode, our host, Joseph, sits down with Brady Perry, Co-founder and CTO of Busie, to discuss the transformation from point-to-point REST APIs to fully decoupled, event-driven architecture. Brady shares the pragmatic “buy vs. build” logic that allowed them to offload Kafka management to Confluent Cloud, freeing his team to focus on solving customer pain points rather than managing infrastructure.
You’ll Learn:
About the Guest:
Brady Perry is the Co-Founder & CTO at Busie, and is a technology entrepreneur focused on solving outdated problems with modern software solutions. A self-taught software developer, Brady brings hands-on experience across TypeScript, JavaScript, React, Python, Django, HTML, and CSS, along with expertise in multiple PLC scripting languages. His background spans both software and operational systems, giving him a practical perspective on building scalable, real-world platforms.
Guest Highlight:
“Data streaming has actually become sort of the centerpiece of the architecture that we’ve built… the stream becomes the API. That’s where the definition is. That’s where the contract is.”
Episode Timestamps:
07:20 – Data Streaming Goodness
19:30 – Beyond the Stream
39:30 – Quick Bytes
45:15 – Joseph’s Top Takeaways
Dive Deeper into Data Streaming:
Links & Resources:
Our Sponsor:
Your data shouldn’t be a problem to manage. It should be your superpower. The Confluent data streaming platform transforms organizations with trustworthy, real-time data that seamlessly spans your entire environment and powers innovation across every use case. Create smarter, deploy faster, and maximize efficiency with a true data streaming platform from the pioneers in data streaming. Learn more at confluent.io.
0:00:00.2 Brady Perry: Data streaming has actually become the centerpiece of the architecture that we've built. Our business is not managing Kafka. Our business should be focused on providing value to our clients. The stream becomes the API. That is where the definition is. That's where the contract is, is the data stream. Because of that, we can move much more quickly with a very small team, and I don't think we would be able to do that if we had never adopted event-driven architecture and Confluent.
0:00:31.7 Joseph Morais: That's Brady Perry, co-founder and CTO at Busie. And this is Life Is But A Stream, the web show for tech leaders who need real-time insights. Today, we're talking about how a small team is modernizing the group travel industry with an event-driven approach. Busie set out to build a marketplace and ended up having to build the market itself. Now they're becoming the centerpiece of group travel booking. We'll dig into how real-time data and streaming let them move fast, stay lean, and focus on customer value, and how the Confluent for Startups program helps support that journey. I'm Joseph Morais, your host and technical champion at Confluent. Let's get started.
0:01:14.3 Joseph Morais: Thanks for coming on the show today, Brady. Let's jump right into it. Tell me about Busie and what do you and your team do, and what does Busie do in general?
0:01:21.5 Brady Perry: Yeah. In a word, the best way to describe Busie is, or in a phrase, we are building the operating system for group mobility. What that means is we're building e-commerce software, operations software, accounting software to help small and medium-sized businesses that run transportation operations specifically for groups of people. So, businesses, weddings, church groups. Major problem around the world is that there is no software facing these bus companies, but they're transporting millions and millions of people a year. So, what we decided to build was something that could help alleviate a lot of the pain points that are facing these companies today.
0:02:07.4 Joseph Morais: Yeah. I live in Scottsdale, and we get a lot of bachelorette, bachelor party buses. I imagine that would be one of your customers.
0:02:16.5 Brady Perry: Perfect use case. Yeah, absolutely. It's funny you mentioned that. My co-founder and I actually discovered this problem on the customer side of things. We were looking to book transportation for a student organization of 150 people when you included guests and we were in college. We had to fax a signed contract to this bus company, and we had to go and hand them a check physically. I don't even know where we found the fax machine, if I'm being honest. I think we had to go off campus to be able to do this. So, we were really confused by that and just didn't understand why we could order groceries to our front doorstep and have them delivered, but we had to go and fax a document. So, what we thought would make sense was kind of like a kayak.com for group travel. But what we learned was, in order for something like kayak.com to exist, there needs to be digital inventory and there needs to be digital pricing, and this industry doesn't have that. And so we flipped that on its head and decided, let's go actually build the infrastructure that can power something like a group marketplace.
0:03:28.8 Joseph Morais: That's interesting. My follow-up question was going to be, who are your customers, but obviously we kind of sussed that out in the first question. So, I think what I'd like to peel the layer back is, I imagine there has to be some regional operator that could be your customer that has built something like this. But where the need was, or the missing element in the market, is there was no easy software for some smaller entity to just kind of onboard and say, "Hey, now I have that digital inventory. I have this web presence, and people can book through a much more modern interface." Is that what you were looking for? And do those kind of one-off systems exist for specific transportation carriers?
0:04:10.3 Brady Perry: Yeah. There are definitely some, we would call them legacy software systems in the industry. They're functionally offline. And what often happens is misquoting, trips just kind of get lost, and then the driver may never show up for their job, or the vehicle just never shows up to pick up the group. This happens all the time. But yes, those one-off systems do exist. Some bigger companies will create their own, but then what happens is it's really hard for them to maintain that. They're not tech teams, they're bus operators or they're transportation providers. So, it becomes too big of an investment, and we're actually seeing a lot of companies today switching off of their in-house built system and coming on board to Busie because of the enhanced interface but also the connectivity to their other systems. We help them to create more efficient routing for their vehicles. That's something that these other systems didn't do. If they were connected, they were connected to Google Maps, and Google Maps doesn't actually account for the vehicle size when routing. So, what we've actually created a routing system that takes the vehicle size into account, takes the commercial aspect of the vehicle into account, and avoids roads that commercial vehicles can't go on.
0:05:27.6 Joseph Morais: All right. So, I really want to get more of those details into what you build in our next segment. But just to kind of put a bow on it, you really have kind of two buckets of customers. You either have customers that aren't running any system at all, and I imagine that is a large number of those smaller enterprises that have been doing these things for years before IT was even a big deal. And then you had these other customers, maybe larger, have these legacy systems that they may or may not maintain in-house. I imagine a large portion of them probably contracted or used a consulting firm, stood the thing up, left them with some runbook, and that was years ago. And those are the customers that already have something, but what Busie has is I imagine light years ahead, better to onboard and has those integration pieces as well.
0:06:12.1 Brady Perry: Yeah. Yeah, absolutely. We are constantly iterating on our product from our customers' feedback, and then just from an onboarding perspective, we can get somebody up and running and comfortable within the system within a single 45-minute meeting and they're off to the races. So, very stark contrast from the systems that exist today where you're looking at months to even a year-long of an implementation process where they're handholding the entire step of the way. They're installing the software system on-prem. That's something that this industry still is dealing with very regularly. So, cloud software, while it was the new thing maybe 15 years ago in the real world, in the bus industry, this is still like new tech.
0:06:58.4 Joseph Morais: Right. They're still figuring out the internet.
0:07:00.3 Brady Perry: Yeah, exactly.
0:07:01.3 Joseph Morais: All right. That's awesome. I can't wait. We're about to transition to our next segment and we'll start to go through all those other details and how they relate to event-driven architecture.
0:07:10.2 Brady Perry: Yeah, absolutely.
0:07:21.2 Joseph Morais: So, we set the stage. Let's dive deeper into the heart of your data streaming journey in our first segment. So, what have you built or currently building with data streaming?
0:07:32.2 Brady Perry: Yeah. Great question. My team is in charge of everything product and engineering-related. We're a very small team, three people today, and data streaming has actually become sort of the centerpiece of the architecture that we've built. So, really, it's just connecting all of our systems together with that nice real-time stream of data so that whenever we add something new, we can tap into that stream, get all of the events that had occurred, build some rich functionality on top of that, and it's super simple for a small team to be able to actually get up and running with that. So, highlight the point that this really is the central piece of our architecture where without it, I don't think we would have been able to build the products that we've built today.
0:08:25.2 Joseph Morais: Just a really quick example...
0:08:26.4 Brady Perry: Yeah, by all means. We do almost have that sort of B2C-facing capability, but it's more white-labeled, so you wouldn't know that you're using Busie under the hood. But one of the things that we provide to our customers, and when I say e-commerce, this is kind of what I mean, we actually can embed rich forms into our customers' websites that capture data and in real time transfer that data to a backend system that our customers are using to manage their sales process and manage their operations, but also allows the customer to get a price in real time on the front end. That's something that just doesn't exist today. But as a traveler or as somebody who's looking to buy something online, you expect to know what the price is going to be as soon as possible, otherwise you might go somewhere else.
0:09:13.0 Brady Perry: So, we're enabling that through event streaming and through just the system that we've built. And then just taking that a little bit, I guess another step further, we also power group bookings for trailways.com. Trailways is a network of independently-owned bus operators. They kind of function like a co-op, but Busie powers the network level or global level of group transportation. So, if you go to trailways.com, you request a quote, we're in real time routing that to the most appropriate operator to service that trip, providing pricing based on their pricing, and then from there that company can manage the sales process. So, sort of maybe like the baby steps of a marketplace, right?
0:10:00.2 Joseph Morais: That's really cool, actually. And I started thinking through that, and you have these companies, plenty of bus companies have been around 20, 30, 40, 50 years and they have probably rudimentary websites with very little functionality. It sounds like you... And previously maybe at best they had an email form. They would submit the number of people they wanted, the dates, they'd wait for an email. And it is crazy to think that we're on the precipice of 2026 and things are still operating that way. But it sounds like you have this ability to embed your rich forms into that legacy website and now suddenly you have this real-time quote ability. When you actually submit and book something, it's actually booking it in their CRM system or whatever equivalent they have. So, even with a legacy system or a legacy website, they can start to get this newfangled functionality pretty quickly out of box, it sounds like.
0:10:51.1 Brady Perry: Yeah, exactly. I think that's one of the benefits to what we provide is you don't need a brand new website. Now, we do offer website services where we will come in and give you a new website and it'll be sleek and it'll be SEO-optimized and all that great stuff, but that's sort of a different product line that is maybe a little bit less interesting, but huge for these customers. They want these modern websites and we came up against that a lot was, "I'd really love to use your product, but I would need to update my website first." So, finally, we were just like, "All right. We're just going to give you a new website."
0:11:26.8 Joseph Morais: That's very smart. I mean, you're taking the market as it comes. You started in this whole marketplace idea, you realized that the underlying systems for these providers were never going to work in that paradigm. Let's build that. And then as you're having these conversations with customers, you realize, Oh wait, they also need their website updated. That's a no-brainer for us to throw that in there to unblock our core tenet of providing this type of technologies. Very cool that you're piecemealing it and figuring out exactly what these providers need.
0:11:57.5 Brady Perry: Yeah. Thank you.
0:12:00.4 Joseph Morais: So, what were the underlying data or technology challenges that led you to adopt data streaming? I know you guys adopted or decided on event-driven architecture early. I'm very curious about how you came to that decision.
0:12:11.8 Brady Perry: Yeah. So, what it came down to was our system was just getting too complex for request-response driven communication between our services. So, what we really wanted was real-time access to data and real-time data flow so that our systems could be actually more loosely coupled between one another. And now all they're relying on is a schema at the end of the day and the Kafka API. But setting up a consumer and ensuring that it follows a specific schema is much better than having or much easier to deal with than having those request-response restful API requests where if system B is down, then system A might not be able to do its job. And that's what we wanted to avoid so that we could have essentially an always-up system, especially for mission-critical pieces of that system.
0:13:08.2 Joseph Morais: So, it's the classic decoupling scenario. That's the bread and butter use case for data streaming, of course. That's always the one I think everyone should start with. It's what it was built for initially and then everything else is just, I think, gravy on top of that when you start thinking... And we'll get more into this in the next segment when you start thinking about stream processing, when you start thinking about integrations to other disparate systems. Definitely the decoupling is the right way to start. Now, I'm going to use your words. Walk me through how you envision data streaming becoming Busie's central nervous system and delivering this cohesive fabric for your customers.
0:13:46.6 Brady Perry: So, I think... It's been a couple years, so I don't remember exactly what happened, but I want to say we were looking into building some, yeah, like real-time alert system based on events that were happening within the system. And we realized, especially with this microservices architecture that we had started to build, that was going to be difficult to do. And so in order to have those real-time alerts driven by real-time events, we identified Kafka being the tool or the technology that would allow us to build that type of functionality into our platform. So, I don't remember where I first noticed it, but just digging in deep to documentation on building these types of systems, it seemed like I kept seeing this Kafka thing over and over again. And I didn't really have experience with... Well, I didn't have any experience with Kafka before that. But learning more about it, it seemed like it was really the right technology to be able to accomplish this type of functionality. I think part of it is also because of the persistence of the events.
0:15:04.2 Brady Perry: When you're dealing with message queues, as soon as the message is consumed, it's gone. Whereas with Kafka, we can set up infinite retention. And that's something that I think is really powerful because, like I said earlier, now we can stand up a new system that is maybe doing some analytics on all of the quotes that have ever occurred within the Busie system. That's something that actually was almost a more recent use case. We built an embedded business intelligence product into our product where we wanted to provide analytics to our customers on the data that we had been capturing for them for years.
0:15:43.5 Brady Perry: That's something that hadn't existed. It's something that definitely doesn't exist in the industry. And we were able to tap into those streams of quotes or trip requests or whatever it is, customers, and actually create some powerful dashboards around that. And all of that was possible because of the way that Kafka works and that infinite retention that I was talking about with our topics. I guess that sort of answers part two of your question of how we started to realize that it was that central nervous system was, after that initial use case, once we started realizing that we could leverage this to build new functionality, it just inherently becomes the centerpiece of the architecture because all of these sort of leaves of your platform are connected to the event broker.
0:16:40.4 Joseph Morais: Right. The persistence piece has always been the most fascinating aspect of Kafka. I've been running it open source, since 2017, and I've always been impressed by the creativity. Obviously, the biggest reason for persistence is asynchronous microservices, right? Not everything consumes at the same pace. In the world of event-driven architecture, you have a lot of fan-out. For every one event created, it's usually consumed at least three times. So, that was the classic implementation of it, or the necessity for it. But I've heard so many different answers of how people use persistence. There's highly regulated industries and they keep infinite retention on because of auditability.
0:17:20.3 Joseph Morais: They can have their auditors come in and just replay the topics for the last seven years and say, "Look, we're good." That's a big boon. But my favorite is the fact that persistence can breed innovation. "Hey, I have this new idea. I don't have to go pull all of my data out of cold storage. I don't have to hook this thing up to a bunch of databases. I don't have to run this bunch of synthetic data. I could just tell it to start at the beginning of the topic." And analytics is a no-brainer. But also in this world of AI, training data. I'm convinced that you could have a business of training ML models just going back into people's old Kafka topics from any given vertical. I am not audacious enough to start that business, though.
[laughter]
0:18:01.8 Brady Perry: No, that's super interesting, though. But yeah, you make a great point. I think there are many answers to that question. But I think that how we started to realize that it was sort of like a central nervous system for us was as we looked to build out new capabilities and we started to realize... I actually think it was the Stream Lineage feature in Confluent Cloud. I know we're going to talk more about Confluent Cloud in a minute, but I just looked at that one day and I was like, "Wow, this has gotten massive." We have so many different streams of data now, all connected to various systems, and it's super interesting. Maybe jumping ahead a little bit, but just thought...
0:18:40.5 Joseph Morais: No, no, no, it's good. Let's tease the audience. I love that response. And yes, we will dive into that quite a bit in our next segment. But first, a quick word from our sponsor.
0:18:57.9 Sponsor: Your data shouldn't be a problem to manage. It should be your superpower. The Confluent data streaming platform transforms organizations with trustworthy, real-time data that seamlessly spans your entire environment and powers innovation across every use case. Create smarter, deploy faster, and maximize efficiency with the true data streaming platform from the pioneers in data streaming.
0:19:34.0 Joseph Morais: Now we'll go beyond the stream on why Confluent is the right fit, or was the right fit for Busie. So, I know that during your initial proof of concept, your MVP, everything was point-to-point and was restful. You realized that things needed to be decoupled. So, you knew that data streaming was the answer. But there's a lot to wrangle with Kafka. What made Busie choose to work with Confluent as their data streaming platform?
0:20:02.0 Brady Perry: I think it was two things. One, exactly what you just said, it's a lot to wrangle. And I recognized that upfront. I was a team of one at that point. I mean, I had my co-founder, but he was not on the technical side of the business. And we wanted to build this functionality into the system, and managing Kafka seemed like something that was going to be out of scope, for sure. So, I think that was sort of one of the impetuses, impeti, whatever the word is.
0:20:31.9 Joseph Morais: I hope it's impeti.
[laughter]
0:20:35.5 Brady Perry: But the second part was, maybe this is by design, Confluent was actually probably my intro into Kafka. The documentation was just so prevalent in anything that I was trying to research as it pertains to Kafka. So, it kind of was like, Here's the things that you need to do, but don't do that. Just use Confluent. And then you can leverage this technology and you don't have to worry about all of these more painstaking processes that are involved in managing open-source Kafka. So, I would say that those are kind of the two things.
0:21:14.4 Joseph Morais: Okay. Well, that perfectly relates into my next question. How did you and your partner choose to embrace a fully managed platform at the earliest stages of building Busie? And I know a lot of people when they get off the ground at the onset, they're like, "Let's just run everything." I know that's the world I came from in the startup that I was in. They're like, "We're not buying anything. We'll spin up an instance and we'll run it." But there's a lot of cost to that. And I know I ran through it and guess what? My startup didn't take off. It's no longer a thing anymore. So, I'm curious what made your team at Busie different that you adopted fully managed services so early?
0:21:51.0 Brady Perry: Yeah. I'd say we are more on the opposite side of that mentality. We are on the buy versus build side of things. The way we see it is our business is not managing Kafka. Our business is not managing databases or EC2 instances. Our business should be focused on providing value to our clients, not those other things. And so we need to focus on creating streamlined communication between dispatchers and drivers. And we can do that with Kafka, but we shouldn't be worried about managing Kafka and ZooKeeper. So, that's kind of where that thought process comes from. Maybe this is naive, but I don't think any company that isn't in the business of managing Kafka should be managing their own Kafka. That's just how I feel about it. Focus on delivering value to your clients. That's the main point. But then also we've gotten so much support from the community, but also from Confluent directly.
0:22:56.8 Brady Perry: The Confluent for Startups program has been amazing in getting us access to subject matter experts. They provide credits, but I think that that's almost... I mean, that's great, but the first part of it is the most valuable where I can talk to experienced solutions engineers or even field CTOs I've had a few conversations with that have pointed me in the right direction when it comes to leveraging some of the newer features that Confluent is releasing, like streaming agents and that type of stuff. So, getting access to those things has also been a major reason why the partnership with Confluent has turned out to be a no-brainer. It's not something that we're going to move away from anytime soon, and very happy that we decided to go this direction to begin with.
0:23:46.0 Joseph Morais: Well, I think you have a very smart mentality because if the intention is to eventually say, "Hey, we're going to just run this for now and then we know we're going to outgrow it and then we'll move later," you then have to pay the pain twice. Because you're learning these ops tasks that you may never use in the future when you get to scale, and then you gotta migrate, which is a pain in the butt too. I am biased. I've worked for Confluent for almost six years now. I ran open-source Kafka and I can still hear PagerDuty going off. So, I'm a little frazzled. I definitely believe, especially... And what you said was key, if you can find a company that you can trust to run the thing. It's different things.
0:24:29.0 Joseph Morais: There's many good companies that are good at running things, and you can trust them, and the total cost of ownership, the value prop is there, I agree, nine out of ten times you should be letting that company run it. Because that's how things scale, especially in a cloud world, you get really good at building one thing. We're never going to help companies coordinate their transportation. That's not what we're good at. We're good at data streaming. You guys do that and let us handle the data streaming. I think it's a fantastic relationship.
0:24:57.3 Brady Perry: You wouldn't have your plumber write software for you.
0:25:00.4 Joseph Morais: Exactly. It's funny, I use that analogy sometimes because people are like, "Well, what do you do?" Because my role is a lot different than it was five years ago when I was on PagerDuty. I used to tell people I was a plumber and I used to turn the wrench, but now I get to talk about the wrenches.
0:25:18.0 Brady Perry: There you go.
0:25:18.4 Joseph Morais: Very apt.
0:25:18.7 Brady Perry: Yeah.
0:25:19.1 Joseph Morais: Now, you mentioned the startup program here at Confluent. And if I recall correctly, last year you were one of the Confluent Startup Challenge runner-ups. Is that right?
0:25:31.0 Brady Perry: Yeah. That's correct. They flew us out to London. I got to speak at Kafka Summit London in front of apparently a couple thousand people. I don't know. The crowd was really dark, but I know I was really nervous.
[laughter]
0:25:46.4 Joseph Morais: I heard you did great, so don't worry about it.
0:25:48.2 Brady Perry: Thanks.
0:25:50.4 Joseph Morais: Okay. I'm glad that... And I really want the audience to know that we're not using the word partnership loosely. There's a real partnership here. There's credits, there's investments from Confluent to see Busie succeed. And I'm glad you've been able to take advantage of those and see your company grow. I'm curious, with the partnership, how did your teams tackle positioning Confluent Cloud for your customer base? Now, I realize that may be completely transparent to your customer base, so if it was transparent, I'm more interested in were there obvious feature enhancements that came for your customers once you adopted event-driven architecture?
0:26:30.4 Brady Perry: Yeah. So, it is very transparent. Unless our customers are paying attention to any blog posts that we've been mentioned in alongside Confluent or anything like that...
0:26:41.7 Joseph Morais: They have no idea.
0:26:42.8 Brady Perry: They would have no idea necessarily the underlying architecture of our system, but what it has provided is our ability to be nimble and very agile when it comes to adding new functionality. And that's because we don't have to go into a monolithic system and figure out how to retrofit this new feature into some very old backend. We can just actually start fresh, build a new system, a new API that's completely standalone. Maybe it connects to a different database that's more purpose-built for what we're actually trying to deliver in this case. Because of that, we can move much more quickly with a very small team, and I don't think we would be able to do that if we had never adopted event-driven architecture and Confluent.
0:27:35.7 Brady Perry: The alternative would be, like I said, dealing with that monolithic system where it's built in a programming language that maybe only one person on the team knows how to write. Now we can hire somebody who... Most, everything that we have is on Node and JavaScript, but if we want to hire a Java developer, they can come on board and start contributing right away because they're going to be building something fresh that doesn't need to be retrofit into an old system. It can just tap into the streams of data or write to the streams of data and do the thing that it needs to do.
0:28:17.7 Joseph Morais: I know I know this, but I've never really thought about it that way, in that you could, because you've decided to decouple your architecture, you could have 10 different microservices running 10 different types of code underneath the hood. It doesn't matter because the common exchange is Kafka. You're going to write an event to a topic, you're going to consume it from a topic. I don't care what libraries you're using in that service. And even though I knew that, I never thought of it as a benefit to a startup or really any type of IT organization for what you just said. Like, Hey, we have this brand new functionality, we know it's going to run better in Rust. Or we know we have a guy that's done this before, he's going to crush it, but he doesn't do JavaScript or Node, he's all about Go. Well, we're going to let him build that in Go because it doesn't matter because he's just going to produce an event from our data streaming platform and we don't care either way. We just want the event to be good and consumed and do what we expect it to do. That's really good. I never thought of it that way.
0:29:15.0 Brady Perry: Yeah. An example of this that's maybe a little bit more concrete that happened more recently. Originally, how we had built notifications was each domain was focused on sending the notifications specific to that domain. So, if there was a quote, something happened in the quotes and bookings world, maybe there was notification infrastructure built into that service that could send the notification, maybe it's an email notification or whatever. But we decided that was going to get really cumbersome as we started to add more functionality, so we decided to centralize the notification system. Now there's one system that delivers notifications based on the event stream, the data that comes from whatever event stream. And that was built in Go, which was something we had never built in Go before, just because we didn't have someone on the team that was really most comfortable in Go. And now we do, and they're like, "It would be great if I could build this in Go." Okay. Go for it. No pun intended.
[laughter]
0:30:17.1 Joseph Morais: That pun should have been intended. [laughter] That's awesome. So, all the goodness. What was the final impact of your customers adopting your platform as their central nervous system in the same way you've adopted Confluent's data stream platform as your central nervous system?
0:30:34.9 Brady Perry: The final outcome, I would say the final outcome hasn't happened yet.
0:30:39.1 Joseph Morais: Okay.
0:30:40.8 Brady Perry: No, but obviously big aspirations for the future of the company, I guess is what I'm trying to say. But I think our customers are pretty happy with our ability to build new things, new cool things that they want and fit those into an existing platform pretty seamlessly because at the end of the day, the stream becomes the API. That is where the definition is, that's where the contract is, is the data stream. And so because we can leverage that, we can move more quickly. Our customers are much more happy because they wanted some report that they were never able to get before and we didn't build for it immediately, but they said, "Hey, it'd be really great if we could see data in this way." And so we can go and do that because we have the data and all we need to do is tap into that data stream and construct the report as an example.
0:31:39.5 Joseph Morais: Honestly, that's the ultimate benefit of a startup. I mean, not that any startup can embrace it, but you guys clearly are, is that you can, especially because of the architecture you started with, you can now listen to customer feedback and realistically implement it. Whereas if you would have started with a monolith batch-based system, it wouldn't have been impossible, but it would be an order of magnitude more difficult and it would have taken longer and you wouldn't be able to see those happy faces when suddenly what was impossible or unknown or just a curiosity becomes an actual piece of functionality. So, could you share some advice and lessons learned for leaders like yourself for tackling data streaming and maybe specifically through the lens of an early-stage startup?
0:32:28.6 Brady Perry: Yeah. I think the advice that I kind of go back to is don't be afraid to ask for help. I'm definitely the type of person who will sometimes feel like if I'm asking for too much, I'm bugging somebody. But definitely what I've learned from working with the Confluent team especially is that they are always open to feedback and they're always, always excited to help you. They will dive in deep with you to help understand your challenges and what you're trying to build. And there's an accountability aspect where I meet with my account team pretty regularly and they're like, "How's this project going?" and "Where can we help?" Lean into that, I guess is what I'm trying to say is probably first and foremost because as a young company, you're going to be a small team, most likely, unless you're lucky or unlucky. But you're going to be a small team and you're going to need all the help you can get. I guess that's kind of the biggest point of advice I would give.
0:33:32.2 Joseph Morais: Yeah. I'm actually really glad to hear that because I've experienced that here at Confluent. I've even experienced that to some smaller degree as a potential customer of Confluent before joining here. And it always warms my heart to hear that because I truly believe it. Plus, what other company has that much experience with all the different permutations of event-driven architecture? That's all we do and that's all of our field does. And whether it's an account executive or solutions engineer or solution architect in our professional services, all we do is talk about event-driven architecture, data streaming, and now a couple of other things like Iceberg and Delta Lake. But that's it. Our scope is pretty narrow. But the benefit to that is if you have any event-driven architecture question, we've seen it a hundred times. And you're right, if you have a tech partner, if you trust them, they can help augment what is generally a small team. So, I'm really glad that your decision to buy has yielded all of this benefit for you.
0:34:36.1 Brady Perry: Yeah, absolutely.
0:34:37.7 Joseph Morais: Now I'm curious about the future. So, what is the vision for data streaming at Busie? What does the future look like? And I know Confluent has an AI accelerator. I'm curious if you're thinking about leveraging that program.
0:34:50.1 Brady Perry: Yeah. I will say that I'm in the camp that AI is going to be most effective when it uses real-time data. I think that that's pretty intuitive for me. So, absolutely, we intend on leveraging those new functionalities that have been added. We actually built three proofs of concept during the AI accelerator which wrapped up last week, Confluent for Startups AI Accelerator, which was awesome, by the way. I know you didn't ask about that, but great program.
0:35:22.4 Joseph Morais: No, I'm so glad. Please tell us all about it.
0:35:24.7 Brady Perry: Yeah. Yeah. I think it's the second year they're doing it or the second cohort...
0:35:30.1 Joseph Morais: I think that's right.
0:35:30.8 Brady Perry: Yeah. Where they select a handful of companies and they show you the ropes to data streaming, but then also gave some deep dives into some of the newer features that have been released lately. So, a deep dive into Tableflow, deep dive into streaming agents. And then there are a bunch of other kind of more businessy-focused things. There were a few sales enablement conversations. I think there was one on building a deck for fundraising. This whole program's created by the Confluent for Startups team and Tim Graczewski, who absolutely killed it for us.
0:36:10.5 Brady Perry: And I've heard great things about the first cohort too. So if you're interested, anybody listening to this, definitely check it out. But at the end, I mean, throughout the 10 weeks, we were tasked with building some functionality on Confluent Cloud using some of these new features if applicable to build an AI, either it's your business or for us it was more of value-add features that we're going to be adding to our product. I can't talk too much about them right now because they're not live yet and we don't want to give too much away. But we're really excited about it. They're going to solve some really big pain points for our customers and I think that's what makes me most excited about it. So, yes, AI is on the horizon. Imminent, I would say.
0:37:01.7 Joseph Morais: Right. To just make it generic, the vision is AI, probably agentic, powered by data streaming to alleviate the pain of your customers.
0:37:12.1 Brady Perry: Yeah, exactly. And I'd say we're never going to be really just like an AI-first company. I don't think we have the right industry for that.
0:37:22.8 Joseph Morais: Sure, I agree with you.
0:37:25.4 Brady Perry: But I do think that agentic AI is potentially going to prove to be most valuable in vertical SaaS, which is, you know, we are a vertical SaaS company, because of the, I guess the dynamic nature of agentic software platforms. I think it's going to be super powerful in vertical SaaS specifically. And so we look at it through the lens of we're going to build a platform that isn't necessarily centered around AI, but leverages AI to again solve those kind of pain points that maybe a rule-based software system is going to be difficult to build. So we can add these AI functionalities on top of that.
0:38:10.9 Joseph Morais: Right. You're building something that's not AI foundational, but it's AI compatible. And I think that is the right way to go 'cause I agree with you, vertical SaaS is a great place for it. But I would even postulate that all industries have some type of dynamic business logic, dynamic data inputs, and that agentic AI is really the solution for that. As much as you want to schematize and standardize everything, and you do that as much as you can, there's always going to be corner cases that have a little bit of ambiguity. And I think that's the difference between a microservice and an agent is an agent could actually handle that ambiguity. And I'm really excited to see where we go with that.
0:38:54.0 Brady Perry: Yeah, absolutely. Absolutely. Stay tuned. I'll keep you in the loop about the cool things that we're going to be building.
0:39:00.7 Joseph Morais: Please do. I'm already thinking, I'm already scheming, maybe like a year out we could have our first ever follow-up episode. We've never done that before.
0:39:07.9 Brady Perry: That'd be cool. I'd be happy to.
0:39:09.2 Joseph Morais: Yeah, especially you get like 12 to 18 months of advancement and it's like a totally new conversation.
0:39:16.6 Brady Perry: I would love to do that.
0:39:27.6 Joseph Morais: So, before we let you go, we're going to do a lightning round. Byte-sized questions, byte-sized answers. That is B-Y-T-E. They're like hot takes but schema-backed and serialized. Brady, are you ready?
0:39:37.2 Brady Perry: I'm ready.
0:39:38.3 Joseph Morais: What's something you hate about IT?
0:39:42.3 Brady Perry: About IT? I think there's still this stigma or stereotype around the people in IT and that they're recluse and antisocial. That's kind of the opposite of what I've found. I've found that a lot of my peers are very easy to talk to, very fun to hang out with. So, I guess it's not really about IT, but kind of about the people looking into IT and what they think about it. That's what I would say.
0:40:08.9 Joseph Morais: I think it's a generational thing for you guys. This is me being an old man for a second, but when I was growing up in the industry, there were a lot more IT curmudgeons. But I do think that is a stereotype that as the average engineer has gotten younger has gone away. So, I love it. I personable engineers.
0:40:28.7 Brady Perry: Yeah.
0:40:29.7 Joseph Morais: What is your hot take on the future of AI?
0:40:35.6 Brady Perry: Okay. So, I don't think that we should be firing people to... Maybe this is not a hot take, actually. That's probably not a hot take. But I think we still have a ways to go, and we shouldn't be firing people. I think the companies that are firing people because they can replace their jobs with AI are going to rue the day that they did that and will come to regret it. Maybe not so much of a hot take, but...
0:41:01.0 Joseph Morais: I think maybe not a hot take, but a very salient, good take. What's a non-tech activity or hobby that's affected the way you think about data?
0:41:10.1 Brady Perry: Gaming, which I guess is still kind of tech, but gaming, like video games and sports. I think it's really cool how starting with sports, now you see all of these analytics things and they're powered by AWS or whatever. Maybe it's a little gimmicky, but there's so much data in sports. There's so much data in gaming. And so when I was learning how to code, I would think about the games that I was playing and how they're capturing the data of whatever action that I'm taking in the video game. Same thing with sports, really cool to see all the data around that. Those are my two.
0:41:49.0 Joseph Morais: Yeah. I really like Thursday Night Football because it's on AWS and that post-game and all the analysis, it's really interesting stuff. But a quick aside, what game are you playing right now?
0:42:00.5 Brady Perry: I mostly play first-person shooters. So, I play Rainbow Six Siege mostly, and then I also play a lot of FIFA, I would say is kind of my go-to as well.
0:42:10.2 Joseph Morais: I'm really into Indiana Jones and the Great Circle.
0:42:13.9 Brady Perry: Oh, yeah?
0:42:14.7 Joseph Morais: Yeah.
0:42:14.9 Brady Perry: Okay. I'll check it out.
0:42:15.8 Joseph Morais: It's first-person. It's not a shooter, but you feel like Indiana Jones. They did an amazing job. Where are you getting outside inspiration from? Is it from a book or a thought leader?
0:42:27.7 Brady Perry: Yeah. I recently finished the book, The Goal by... I'm going to maybe butcher the name, Eliyahu Goldratt. Have you ever heard of this? Yeah. Great book, like business novel. Really enjoyed that. And as a founder who, while I'm on the tech side, I have to wear many hats, I found that it was pretty eye-opening to how we were running our business and definitely was able to draw some insights from there. But then now I'm reading a book called Good to Great by Jim Collins, another kind of classic business book.
0:43:02.3 Joseph Morais: That one I've heard of.
0:43:03.5 Brady Perry: Yeah. Really liking it so far. I'm not so far into it yet. I just finished the other one. Those are kind of the two main sources of inspiration. I mean, I guess I get a lot of inspiration from a lot of different places and people. Get inspiration from my team, from my family. But yeah, I'd say book-wise, those are kind of my go-to's right now. I also got a signed copy of Kai Waehner's Ultimate Guide to Data Streaming, which I'm really excited to dive into.
0:43:31.6 Joseph Morais: I unpacked that book for him at [0:43:33.1] ____.
0:43:33.4 Brady Perry: Oh, really?
0:43:34.0 Joseph Morais: Yeah.
0:43:34.3 Brady Perry: Then we must have met. We must have met in person.
0:43:39.0 Joseph Morais: Yeah. I was just the tall guy in the background just unpacking it while Kai had that huge line. But when we have a debrief, I'm going to tell Tim G. You gave... He was the thought leader answer for that question.
0:43:50.9 Brady Perry: Yes. Okay. Yeah. Tim G. Also. It was Tim G.
0:43:55.6 Joseph Morais: So Brady, any final thoughts or anything to plug?
0:44:00.8 Brady Perry: Connect with me on LinkedIn. I think it's Brady-Perry at the end of my URL. Want to shout out Tim G for real. This was actually part of my plan. Tim Graczewski has been amazing. Raquel Sorensen, my AE, account executive, she's been awesome, super helpful. I think both of those people actually set up this conversation, so I'm really appreciative to them for that. And then I'll also shout out my friends Tim Wade and Saul Sparber. Tim Wade is one of the founders of TwinLabs, and then Saul Sparber, the founder of Agent Taskflow. Two guys I actually became friends with through Confluent, believe it or not, through the Confluent for Startups program. Love those guys. They're great guys, great companies. Check out their companies. And yeah, I think that's all I got.
0:44:52.9 Joseph Morais: That is a very useful platform or very useful platform for a call out. So thank you for that, Brady. So, thanks so much for joining me today. And for the audience, stick around because after this I'm giving you my top three takeaways in two minutes.
0:45:17.5 Joseph Morais: That was a great conversation with Brady. Here are my three top takeaways. The first one's gotta be this idea that Brady and his founder wanted to build this Kayak for the group travel space. But when they started to peel back the layers and they started to reach out to the providers, they realized, wait, no one even has the software or the data to do this. So, they pivoted and they said, You know what? Instead of building the marketplace, let's build the market. And I think that's a really smart opportunity for a startup that said, Hey, we saw this disparity in the way we book group travel compared to anything else we're doing in our lives in, I guess this was 2020, 2021, 2022. And they saw this opportunity, said, "Let's seize this opportunity," realized that the opportunity was actually more foundational than that, and then built that on top of event-driven architecture, which will ultimately allow them to build that Kayak service if they ever choose to go B2C.
0:46:14.8 Joseph Morais: The other takeaway I really liked, and this is for any Confluent customer, but in particular for startups, do not be afraid to ask for help. If you're an early-stage startup like Busie was, small team, right? Ask the experts for help. Allow them to augment and supplement your workforce by one, spending the time to answer your questions or to give you that code snippet that's going to get you unblocked and not have to chase around the answer. Again, our folks here at Confluent, this is all we do: Event-driven architecture, data streaming, stream processing, et cetera. We have the answer, so do not be afraid to ask for the help.
0:46:55.2 Joseph Morais: And then future thinking, my last takeaway is Brady's assertion that AI will be most effective when using real-time data. I agree. Obviously, I'm a bit biased, but real-time data is the best data. The only thing newer than real-time data is future data, and I joke about this a lot, but that is just not possible. Maybe with quantum computing it will be. And the fact that he thinks that that is the right direction to go with real-time data, I love that he was able to be part, or Busie was able to be part of the startups accelerator. So, they got exposed to Confluent Intelligence, which is our newest suite of features that make building agentic AI on top of streaming much easier. So, I just think that this whole story of trying to build one business, pivoting to another, getting help from Confluent in various ways really accelerated the success of Busie, and I can't wait to see where they are in a year or a year and a half from now.
0:47:54.8 Joseph Morais: That's it for this episode of Life Is But A Stream. Thanks again to Brady for joining us, and thanks to you for tuning in. As always, we're brought to you by Confluent. The Confluent data streaming platform is the data advantage every organization needs to innovate today and win tomorrow. Your unified platform to stream, connect, process, and govern your data starts at confluent.io. If you'd like to connect, find me on LinkedIn. Tell a friend or coworker about us and subscribe to the show so you never miss an episode. We'll see you next time.