Brendan Kohler CTO of Sentenai

(Joel Beasley at 00:00:00) Today, we are talking to Brendan, the CTO of Sentenai, and we discuss having a diverse set of viewpoints within your company, building a recruitment process that is self-sustaining, and saving time by keeping your development processes in-house. All of this right here, right now on the Modern CTO podcast. Here we go. This is the Modern CTO podcast. Yeah.

(Joel Beasley at 00:00:34) I actually wanna know. I read a lot of biographies too.

(Brendan at 00:00:37) Oh, yeah? Yeah. It's funny. My parents, you know, they have no idea what to get me these days. So a lot of things about famous engineers and stuff like that.

(Joel Beasley at 00:00:52) You get the Elon Musk one?

(Brendan at 00:00:55) I have not gotten that one.

(Joel Beasley at 00:00:57) Oh, I highly suggest it.

(Brendan at 00:01:00) I'll have to put it on the list. These days, I don't get too much time to catch up on my backlog.

(Joel Beasley at 00:01:08) What? You roll with Audible?

(Brendan at 00:01:12) I really don't. I prefer the printed books.

(Joel Beasley at 00:01:17) But it gives you that ability to hear the—you know, you just you don't have to be sitting down and reading.

(Brendan at 00:01:24) Yep. So like in the car.

(Joel Beasley at 00:01:25) True.

(Brendan at 00:01:26) You know what the problem is? I find I have a lot of trouble on the concentration side. I'll just wind up doing something on my phone.

(Joel Beasley at 00:01:38) Like Candy Crush?

(Brendan at 00:01:40) Yeah. More like checking my email, looking at GitHub issues. It's the constant worry. You know, what's going on in the organization and can I stop something bad from happening?

(Joel Beasley at 00:01:55) Okay, Batman. So what do you think of Slack? I was reading an article the other day. This guy was talking about how it's connected us like never before, but it's stressing out a lot of engineers and creating a lot of productivity issues with the on-demand access to everyone.

(Brendan at 00:02:14) Yeah. It's interesting. I had to get rid of a lot of my Slack groups, but internal Slack doesn't seem to be much of a problem. We actually use it so little. Yeah.

(Brendan at 00:02:28) Yeah. We use it to communicate, but we don't have 250 channels. So it hasn't been much of a distraction for people. We actually use it a lot for file transfer.

(Joel Beasley at 00:02:46) Yep. We do too. Yeah. Copy and pasting to the team, and file transfer is one of the biggest uses of Slack for us.

(Brendan at 00:02:55) Yeah. Yeah. We don't have—we have a few channels around things like sports or music, but it's one or two posts a day.

(Joel Beasley at 00:03:06) Oh, what type of music you like?

(Brendan at 00:03:09) Oh, I'm all over the place. Lately, I've been really into Scandinavian black metal.

(Joel Beasley at 00:03:19) Alright. So for the interview, I'm gonna write down Taylor Swift. Oh, man. My producer guy, he just did the rock symbol. He was just talking about Scandinavian black metal like three days ago.

(Brendan at 00:03:33) Oh, that's great. I have a lot of friends who are musicians. We actually have a Slack among all my childhood friends where we just keep each other kind of up to date. And we exchange music a lot. A couple of them are musicians, and I think they play so much music, they only get interested in really, really weird things.

(Joel Beasley at 00:03:57) Yeah. You know what? I actually have a friend, Ryan, and he's a big-time musician. I'm a musician as well, right?

(Joel Beasley at 00:04:04) But he's getting really into it. So then he started to get really off into this strange sort of funk mix stuff. And it's cool because it's technically correct, but when you listen to it, it's like, this is not enjoyable.

(Brendan at 00:04:18) Yeah. Half the YouTube videos on our friends music channel, it's all stuff with 5,000 views and hard to listen to.

(Joel Beasley at 00:04:33) That's a Friday night for you, right? It's like talking to my wife about single responsibility principles. Tell me more.

(Joel Beasley at 00:04:45) So I saw that you had this data science query engine. And right when I saw that, I was so excited to talk to this guy because being able to collect all that sensor data and filter through it rapidly—the task should not be learning how to write the complex query. The task should be going in and just getting the information you need.

(Brendan at 00:05:06) That's absolutely true. That's a very succinct way to put it. We have trouble to this day, you know, a few years in with messaging with organizations sometimes because a lot of them don't realize that it is a problem because they don't actually ask their data scientists.

(Joel Beasley at 00:05:26) Yeah. Well, what's the conversation to have, Brendan? Hey. There's a tool that replaces your job. Should we implement it?

(Brendan at 00:05:35) Yeah. Right? Well, we actually tried that with data engineering teams embedded in IT departments, and we found they really didn't want to implement a tool to replace their job. It was—they didn't even like parts of their job, but our messaging was so much about automation of a manual job that I think it scared a lot of people. But now we went and talked to data scientists, and they were like, oh, that's—you know, my biggest problem is IT has this terrible system they implemented to store the data. They give me a weekly dump of SQL Server data. Or they have an internal API on S3 buckets, and we have to parse it all ourselves. So we have shadow infrastructure in the cloud.

(Joel Beasley at 00:06:25) And no docs, by the way. You can just get it off the HTTP request. You can just guess and check, right? You just look at the logs, figure it out.

(Brendan at 00:06:33) Yeah. I think people have in most industries come to realize the value of their data for future success as a business, but it's hard kind of crossing that chasm between that realization and saying, like, we're going to take specific steps to improve our business processes using data or predict and prevent failure. It's such a mental leap for them that there's a lot of education involved.

(Joel Beasley at 00:07:07) Yeah. So have you tried the messaging of we're taking care of the boring stuff so you can build cooler stuff on top of it?

(Brendan at 00:07:16) We definitely have, and it definitely speaks to data scientists. I think it's a good kind of messaging for data scientists. For business owners, we're finding that they're most concerned about failure, not about boring stuff.

(Joel Beasley at 00:07:35) Failure in how?

(Brendan at 00:07:37) Failure of data science projects. So they're worried if they hire a data science team to do this big initiative around becoming data-driven, that initiative will fail because they're not prepared for it, and therefore, they will have wasted a lot of time and money.

(Joel Beasley at 00:07:57) Yeah. So then that becomes a sales thing, right? So now it's up to you to say—now it's like you have to say, look. You didn't get to where you are by being scared to try new things. You got to where you are by expanding and trying new things and growing. This is not the time to recoil back and be scared of something new. This is the time to say, hey. There's something new. This is an opportunity. We have to try it or our competitors are gonna pass us by. This is something that needs to be done.

(Brendan at 00:08:26) Yeah. Absolutely. And actually, we found the best positioning to be that we make enterprises data science ready. And basically, the idea that just by implementing us as a repository for all of their sensors and metric data, that they drastically reduce the risk and can hire a data science team without fear that they won't even be able to get started.

(Joel Beasley at 00:08:55) Oh, so you're—ah, I love it. So that's a brilliant way to position the product, right? You could amplify their fears now. Dude, you're gonna fail. This is—you gotta fail if you don't train, man. You need a training program. This is what we do. Here you go.

(Brendan at 00:09:11) I know. All our slide decks are about fear now.

(Joel Beasley at 00:09:16) Media company. Oh, man. Alright. So is this what you're most excited about right now, this data science query engine? Is that where you're spending most of your day?

(Brendan at 00:09:31) Yeah. So it definitely is. I mean, I think actually, I think it's a little higher level than that. It's a means to an end. But, you know, I've spent the past ten years going between research and industry in industrial machine learning. And there's this, I think, general concept that has had a lot of different names over the years—from complex event processing to digital twin to sensor fusion—that really encapsulates the kind of value that can be created with sensor-driven automation. And, you know, enabling that in the future is really where we're trying to put all of our efforts in. So, you know, even looking at the R&D efforts of data scientists, that's a stepping stone to enabling full autonomy both in the factory or in city optimization or in logistics networks.

(Joel Beasley at 00:10:33) That sounds awesome. I'm geeking out over here. I'm listening to you talk about this while looking at your site about all the example code of, like, turbine equals false if weather's cloudy. I'm like, oh, man. Why does this make me so happy?

(Brendan at 00:10:51) No. It's funny. I started out actually doing this in research for government applications where we would have telecommunications data and a bunch of other kinds of data, and we would have to synthesize all of that into intelligence that could be used to essentially prevent disasters, usually regarding existential threats in foreign countries. And the amount of data that needed to be synthesized and run, or put into the simulations in the supercomputer clusters was just so massive that it changed my outlook on what a really interesting world-changing problem is to solve. And I went from thinking about peer-to-peer file sharing all of the time and how to do that better to, like, we can solve massive problems if we can better combine different models, different sensor data, different intelligence to actually be better at predicting the future. And, you know, effectively in doing that, reduce the dimensionality to increase the fidelity of our predictions. And so I've sort of, for the past ten years, kind of—it's become my mission.

(Joel Beasley at 00:12:11) So are you working—it's like a lot of the companies, you know, when cloud came out, everybody got real—all the marketing departments went crazy, right? Cloud. We're gonna say everything. The cloud. And then you get the machine learning and AI, and everyone's all pumped up about that. Well, I've seen a lot of the IoT and the enterprise stuff with IBM, some different companies like that. They're like, we're the IoT cloud. How do you guys—how do you see that?

(Brendan at 00:12:36) You know, it's for us, it's actually useful because it's a very partner-driven environment. And all of those IoT cloud solutions are really focused around asset management and connectivity. So they actually don't have a very good story for how you go from raw data to actionable insight. And so we actually utilize all of that marketing they put out there to figure out who to go after for partner relationships.

(Joel Beasley at 00:13:08) Do you have a partnership with IBM?

(Brendan at 00:13:10) No. Not IBM.

(Joel Beasley at 00:13:12) Oh, I'm friends with one of their people that does this at IBM's IoT dev cloud division if you want me to introduce you.

(Brendan at 00:13:19) Oh, that'd be cool. That's in Germany, right?

(Joel Beasley at 00:13:22) I am not sure, but I talked to her on LinkedIn. I was commenting one day on the mayor or the chief technology officer of New York City, right?

(Brendan at 00:13:36) Mhmm.

(Joel Beasley at 00:13:37) And I was like, hey. What's up? We were just talking about some geeky thing, right? And we're going back and forth, and I click on his name. I'm like, oh, dude. He's the CTO of New York City. That's kinda cool. We shot him on the show. So then this other girl chimed in to our conversation, and I was like, oh, who's this? And it was—she was some head of something in the IBM IoT. So I said, oh, she looks really cool. I was reading about her past. I was like, she's really smart. So I said, hey. You wanna do a fifteen-minute call? And so we jumped on a call, and I was talking to her. And she was really sharp, and they do all sorts of stuff with—there's like a—it's like a small team. You know, like Alexa and all these teams, they're not as big as you think.

(Brendan at 00:14:19) Oh, yeah. Absolutely. Yeah. It's actually kind of amazing how small the teams in practice are around what are big marketing initiatives.

(Joel Beasley at 00:14:30) Right? Like, we talked to the Microsoft team that does their devices, like their HoloLens and their Cortanas and everything like that.

(Brendan at 00:14:37) Yeah.

(Joel Beasley at 00:14:38) Super smart. And they transmit information, and they move really fast. And I'm just like, it's cool to interact with them.

(Brendan at 00:14:46) Yeah. I think it's actually—

(Joel Beasley at 00:14:48) Because I—

(Brendan at 00:14:49) Oh, go ahead.

(Joel Beasley at 00:14:50) No. It's just that previously, I'd been interacting with companies where I just talk directly to the CTO, right? Or I talk directly to the owner, right? And because it's usually startup or mid-level. But then when I started working with these really large organizations, I saw how they kind of quarantine their little projects, but there's also communication between them into these essentially—it looks like a distributed cluster. And then I thought to myself, I'm like, you know, this is so smart, right? Because it looks exactly like how if you were to make a visual representation of an application I wrote. It's like, I guess this is the way to scale, law of Demeter style, right? It's smart.

(Brendan at 00:15:28) I think it's totally true. And if you look at, you know, kind of some of the classic management literature, there's really only so many people that you can manage leading a particular product. And I think the successful projects that we see are really, you know, small and focused because of that, and they've externalized all of the functions that are not core to what their actual process is.

(Joel Beasley at 00:15:58) Exactly. It's like when you're watching the Steve Jobs movie. The people in the room actually making the core device, there's like five, six.

(Brendan at 00:16:06) Yeah. Absolutely.

(Joel Beasley at 00:16:08) So I wanna know about—you've got—Spanish word for size is what?

(Brendan at 00:16:13) For size?

(Joel Beasley at 00:16:16) Yeah. Talla.

(Brendan at 00:16:18) Oh, is that true?

(Joel Beasley at 00:16:20) Did you not know that?

(Brendan at 00:16:21) I did not know that.

(Joel Beasley at 00:16:22) Yeah, dude. Oh, man. I'm sorry. I thought you had this whole sweet story about size and teamwork and knowledge base, and that's why you called it Talla. I was giving you the setup, buddy.

(Brendan at 00:16:33) Oh, man. So that's actually funny. We—that's kinda funny because I see no connection to Talla in that. But you know what it is. So I just assume everything's made up because when we named Sentenai, we spent like three weeks looking for a .com that was pronounceable. And so we were just making up word after word. So if somebody has the .com, I just assume they made it up now.

(Joel Beasley at 00:17:04) So that's the story of Talla.

(Brendan at 00:17:08) Well, Tala is actually an interesting story because Tala is kind of the story of what could be right and what is wrong with the investor ecosystem in Boston. Yeah, I've been here for quite a while, and I cofounded the venture fund Hyperplane because it's such a hard thing to raise capital in this city to solve hard problems where you're effectively doing kind of a moonshot effort. And we met Talla's founder and CEO, Rob May, kind of in the process in a coworking space when he was starting up, and he had just exited his previous company as it had just been acquired. And he was so much like the founders we wanted to bet on that we just essentially, you know, right there made the decision that after like two meetings, we're gonna bet on this guy.

(Brendan at 00:18:15) You know, none of this classic Boston stuff about needing a business plan or, you know, a use of proceeds. We're just gonna, you know, make the bet right then. And we had actually a lot of success in our first fund making bets like that. And I wish more venture funds in Boston would kind of be more risk-taking like in the Valley.

(Joel Beasley at 00:18:36) Well, you know, that's one of the reasons, that's one of the excuses to actually have capital like that. If you have a business that's straightforward, here's your customer. If you provide the solution, they will buy it, then you can just go directly to that person and get them to jump on board your project and sell them and then use that cash to actually produce your product. And you can essentially grow without a huge capital injection or possibly no capital at all. But when you're doing something like science or when you're doing something that's like crazy moonshot, historically, that's when you go get the capital.

(Brendan at 00:19:15) Yeah, absolutely. And I feel like investment firms, I think become more conservative over time. You know, as they grow, as they have more successes, more wins, they don't want to have a loss because they may not raise another fund after that. So I think they've stagnated.

(Joel Beasley at 00:19:37) Yes. Like the attorney that doesn't wanna take the case because he wants his or her 99% win rate. You know?

(Brendan at 00:19:42) I think it's exactly the same thing. That's also why many funds move upstream as well. As they grow, they'll start investing B's or C's because they don't wanna invest in riskier deals anymore.

(Joel Beasley at 00:19:55) Valid point. Are you actually in Massachusetts?

(Brendan at 00:20:01) I am.

(Joel Beasley at 00:20:02) Oh, nice. So you're on premise with your whole team?

(Brendan at 00:20:07) Yes. I am.

(Joel Beasley at 00:20:07) And that's working out pretty well, right?

(Brendan at 00:20:09) It is. It works out amazingly well for us. It to the point where we were looking at hiring some remote workers, and our employees asked us, our engineers specifically asked us not to. They prefer working face to face, whiteboarding, doing all that.

(Joel Beasley at 00:20:27) The amount of time you save in the communication of being in the same room as the person when you're communicating an idea, especially when you're doing something that's moonshot style or hasn't been done, is so valuable. You save so much time by having everyone there.

(Brendan at 00:20:44) I think that's right. And it matches kind of my experience. And there was a time when I lived in Boston and commuted to Connecticut to do distributed sensor networks research at Yale. And my experience was always that, you know, despite working harder on my remote days, I would get a heck of a lot more done on my on-site days.

(Joel Beasley at 00:21:11) Absolutely. So have you been advising startups recently?

(Brendan at 00:21:15) Yeah. So after the—so I was part of a research lab at Yale, and we spun out a couple companies from that lab. One of them got acquired. And as soon as we got acquired, we were acquired by this, basically, this energy company that became a quite a, you know, nine to five job post-acquisition. So I started mentoring startups at the Harvard Innovation Lab and at various places like Techstars and Lean Startup Machine and such.

(Brendan at 00:21:49) You know, it basically is a way to pass the time. And I've tried to continue that, especially—I teach entrepreneurship classes to students who are interested in entrepreneurship because in the late nineties, I started my first Internet business, and there was really no support for a teenager trying to start a business of that kind back then. And so I've always felt strongly about, you know, giving back because it's not impossible for a teenager to create a good online business or software product. It's just that the material, the information, you know, and insight hasn't been out there. So the nonprofit that I teach for is very focused on enabling people to get started with entrepreneurship earlier in their lives.

(Joel Beasley at 00:22:40) Love it. I was starting when I was 13 years old, and it was incredibly difficult. I was so happy. I'm early thirties. I was so happy when PayPal came out because I could do business online.

(Brendan at 00:22:54) Me too. Me too. Yeah. PayPal was there. Thank you, Yuan.

(Brendan at 00:22:59) Yep. It was.

(Joel Beasley at 00:23:01) Have you seen his flamethrower?

(Brendan at 00:23:05) I have. And I was very tempted to buy one, but my wife wants to buy a house. So I think I'll be putting everything towards down payments rather than a $500 torch.

(Joel Beasley at 00:23:19) What are you really excited about right now today? Like, what are you pumped up about that's happening at the organization?

(Brendan at 00:23:25) Yeah. I think there's two things. We just hired our first sales director, and it's really changing how we think about product marketing and enterprise sales. And the fact that it's not the, you know, founders that are out there doing a lot of the sales now, or that we're taking a back seat and letting an expert drive, it's really exciting to see that evolution in the organization. So that's nice.

(Brendan at 00:23:50) And then the second thing is I'm really excited about the challenge of building a highly skilled functional programming team around, you know, distributed systems and machine learning because it's an area that—you know, it's always tough to recruit functional programming people because there's a lot fewer of them. And it's always tough to recruit people who are experts in distributed databases or machine learning. And so the challenge of actually combining that and kind of having people lean on each other's strengths because there's no unicorn out there who's an expert in all of it is a really interesting one, and it's exciting to kind of build a small team that's highly focused.

(Joel Beasley at 00:24:42) So how are you filling your pipeline with potential candidates? Like, how do you keep people in front of you, and how do you make the decision about who to pull on?

(Brendan at 00:24:50) Yeah. So one thing that we do is we get everybody on the team involved. You know, it's sort of recruit your own coworkers kind of thing. And everybody's excited about what we do. Practically all of our engineers actually moved to Boston to join the team.

(Brendan at 00:25:12) So, you know, they're not afraid to go and recruit people from elsewhere. But we also spend a lot of time—I think there are specific areas where, you know, in recruiting, you can kind of have a high impact for less work. And I would say the AngelList recruiting has been a godsend for that because there's a lot less noise than, you know, all the other job boards and job post things. You don't get, you know, you don't put an ad saying specific technology required and get 50 people who have five years of enterprise Java experience that has nothing to do with what you posted applying. And so that's been a huge time saver.

(Brendan at 00:26:00) But the other thing is we go where the functional programmers are. So we go to the meetups. We go to conferences. You know, we think about sending the whole team to Strange Loop. You know, we do things, post specifically where we find functional programmers hang out.

(Brendan at 00:26:19) And, you know, we've even sponsored podcasts that we knew functional programmers will listen to, and at least one person on our team joined because he thought that meant we must be a legit company.

(Joel Beasley at 00:26:32) That's what it means, though.

(Brendan at 00:26:35) Yeah. So it's, I think it's such an ensemble of, you know, methods that you have to employ when there are just relatively fewer people out there. But the baseline of how experienced and how knowledgeable the candidates we get are is way above my last company when we were mainly, you know, trying to recruit general Python engineers and having to kind of wade through a bunch of people who, you know, barely knew Python but knew it enough to say they'd apply for it in a job.

(Joel Beasley at 00:27:20) Yeah. You want people that like, know it.

(Brendan at 00:27:24) Yeah. That's absolutely it. So we found our magic formula is just saying that we're doing distributed databases, machine learning, and Haskell, and people just kind of apply.

(Joel Beasley at 00:27:36) You ever think about recruiting authors of books on the subject?

(Brendan at 00:27:41) We actually—our first employee was an author. So we thought about that early. Like, recruit people who lead meetups. Like, one of our engineers is co-organizer of the local Elm meetup. So it helps kind of getting a pipeline.

(Brendan at 00:28:02) You know, we don't hire that fast because we're a believer in small teams, but that does give us the opportunity to be really picky about who we hire.

(Joel Beasley at 00:28:10) So I hire fast, but on a, like, a come hang out with us for a week or two and see how it goes basis. Right? Like, I want people to come hang out and see how they work because you can't tell. Even if you have three interviews for an hour, you really can't tell that person the same way you can as if you just work with them in the office for a week.

(Brendan at 00:28:32) Oh, yeah. Especially if you're trying to do like whiteboarding sessions or code interviews. It's just hard. And I think the, yeah, the challenge when it comes to these more niche skill sets is that you really wind up having to fly a lot of people into the office to spend a day. And we've done that, you know, with a number of people.

(Brendan at 00:28:54) But we also—we've had one of our engineers actually asked us if he could just come try working with us for, I think it was three weeks that he wanted to come. And we thought, why not pay for it? So it actually, I think it's worked well to have people come hang out. We've gotten a much better sense than in technical interviews.

(Joel Beasley at 00:29:18) Yeah. Well, you get the person that really wants to be there too, especially if you're getting people calling you up saying, hey. I wanna come work for you for like three weeks. I just wanna show you what I can do. Man, when I get that call, like, I'm going to be so happy.

(Joel Beasley at 00:29:32) Right? Like, I'm putting it out there because those people aren't hungry. I got people that won't even show up for job interviews. I'm like, oh, man. Come on. You know?

(Brendan at 00:29:41) Wow. I haven't had that one happen, but I can imagine.

(Joel Beasley at 00:29:46) Not on the programming side.

(Brendan at 00:29:48) Oh, yeah. Oh, not on the programming side. Okay.

(Joel Beasley at 00:29:50) Yeah. Not on the programming side.

(Brendan at 00:29:52) Oh, yeah. We've definitely had sales interviews with people blowing us off. So.

(Joel Beasley at 00:29:58) Yep.

(Brendan at 00:30:00) Yeah. So, you know, someday I hope to have kind of built a recruiting process that is self-sustaining so that I'm not at the forefront of it every day spending an hour on it. But I still feel like we're at that stage where every new person changes the culture in the company and so that we are trying to be careful also, you know, really focus on having a diverse set of viewpoints in the company. Um, you know, it's so easy to hire a bunch of people that—it's like, well, they're like me, so I know they're gonna do well.

(Joel Beasley at 00:30:41) Well, I'm always trying to find my clones.

(Brendan at 00:30:46) Yeah. Yeah. So that's what I did at first, and now we've been trying to find people—I would say find people who make me slightly uncomfortable from a, like, skill set standpoint. Like, they're either taking on skills that I feel uncomfortable with and they've, like, embraced learning those skills in addition to programming, or they, you know, are experts in something that I actually, you know, dislike or maybe discount the importance of. But, you know, intuitively or, you know, in theory, know it's important even if I hate it.

(Joel Beasley at 00:31:27) Yeah. It's always good to find that person that like, just knows a little bit more than you too.

(Brendan at 00:31:34) Yeah. Oh, no. Absolutely. That's the one constant interview question. I always try to—and even in the first call—try to find that one thing that somebody, you know, especially if it's something in relation to what we do because I have some level of expertise in all of the areas that we work in.

(Brendan at 00:31:59) I want them to, like, go in that part where I think they must know the most, like, demonstrate how much more they know than me.

(Joel Beasley at 00:32:09) Yeah. That's how you go through that dance of validating an expert.

(Brendan at 00:32:13) Yeah. Absolutely. And I think we try to not have too much overlap in expertise. Like, there's a general baseline we look for, but, you know, we have people with twenty years of experience on kind of like being deep in the ops community and people who are category theory experts and mathematicians. And getting those expertises together has yielded for us in practice—I don't know if we even, you know, really realized it at the time, but it's yielded some amazing insights from kind of like cross-field pollination.

(Joel Beasley at 00:32:53) And how big is the team currently?

(Brendan at 00:32:55) So we're 12 people right now.

(Joel Beasley at 00:32:58) And how many of those are engineers?

(Brendan at 00:33:01) All but three. So it's, yeah. And we have one person who doesn't code. So.

(Joel Beasley at 00:33:13) Who's that?

(Brendan at 00:33:15) Our director of sales.

(Joel Beasley at 00:33:19) Alright. Well, that could slide because the people that bring money in, they don't have to necessarily code.

(Brendan at 00:33:25) Yeah. Absolutely. It's funny because our director of sales actually has experience in selling technology products that are kind of like ours. Maybe not the same kind of thing, but, you know, similar in complexity to sell into an IT organization or into a business unit. And so that's been great.

(Brendan at 00:33:47) And the rest of us, you know, even my cofounder, our CEO, he was an MIT CS major, and he kind of went out of CS into venture and, you know, that route with his life. So he doesn't code much anymore. But I always laugh at the fact that literally everybody on our team has a, you know, CS or CS-like degree except for our sales director.

(Joel Beasley at 00:34:17) So I have a question. I've gotta know this one. If Elon Musk like texts you today, right, is like, hey, Brendan. Like, come over to my house and hang out. The first question is, would you go?

(Brendan at 00:34:31) Yes.

(Joel Beasley at 00:34:34) Alright. Now you're at his house. You're hanging out with Elon Musk. You ask to shoot the flamethrower. He says no. Sorry about that. So and then you bring Cybertruck up and he shuts you down on that, right? No. But he's got a time machine, right? And you get to go into the time machine and talk to yourself ten years ago and give yourself one piece of advice, what would it be?

(Brendan at 00:34:57) Oh, man. If I had to give myself—so ten years ago, it would be not doing research is the right decision. Entrepreneurship is the right thing. And be very, very, very careful about which projects you work on and companies you join. Optimize for how quickly you can personally learn, not for safety or for, you know, personal relationships.

(Joel Beasley at 00:35:29) I love it, man. That's great advice. Your previous self thanks you, by the way.

(Brendan at 00:35:34) It took me way too long to learn that, but I think it's an important lesson. And I see it in entrepreneurs. Back when I was doing venture full time, I saw it in a lot of entrepreneurs that they were not afraid to kind of take those risks and really put themselves out there as a first-time entrepreneur. And those are always the ones that I wanted to bet on and the ones who I think in the end have been the ones who have been driven to succeed.

(Joel Beasley at 00:36:07) Yeah. Some people, they just have that fire in them where they just get up and go, go, go, and nothing in the world could possibly slow them down.

(Brendan at 00:36:13) Yeah. My filter now is, do I want to or would I ever bet against you?

(Joel Beasley at 00:36:21) Oh my god. Dude, you need some Tony Robbins, Brendan. You need some positive in your life, my friend.

(Brendan at 00:36:31) Yeah. I like survivorship, but I think there's some kind of irrational exuberance around fundraising and fundraising as a startup milestone. And I think it's the companies and the founders who actually are kind of shaped by the kinds of ordeals they have to go through, and they don't have that easy path when they're not fresh out of school from a top university and somebody gives them, hands them three or $4 million and says go build something cool. I—it's the opposite. The people that I've met who have been people who are starting companies after five or six different tours of duty in that space, who have really interesting life experiences.

(Brendan at 00:37:39) One of my cofounders of the venture fund that I cofounded, he had come to America with nothing for medical treatment, and he built his whole life up from scratch from that. And those people have such an interesting outlook. I can never fully relate to it myself, but they just have such an interesting outlook on life, and they've learned positivity. I think that, to me, that's what makes them survive when other people kind of decide to give up and go join Google.

(Joel Beasley at 00:38:18) Yeah. So when I did this book, I had the editor kind of interview me, asking me questions while we were finishing the book up. And they started asking some really deep personal stuff because they wanted to dig in for some reason. I'm like, alright. So they're asking me, why—because to me, I'm mean. Right? You know? But they're asking, like, why don't you stop? Like, what makes you keep going? Like, why are you the way you are? And I'm like, dude, I don't know. And so they were asking me questions about childhood. And then the ultimate thing that they came to was when I was 13, I got hit by a car, and I was in a wheelchair for a year, and I had to learn how to walk again. And then halfway through my rehab, I fell and re-broke my legs and had to learn how to walk all over again. And that was this big, big, long two-, three-year chunk of my life.

(Joel Beasley at 00:39:11) And then they're like, well, now you have the hips and legs of an 80-year-old woman, and you're not gonna be able to do sports. And then I ended up joining the football team. Right? And then I just kept pushing forward. And then I'm like, you know what? Maybe that's where I learned how to be positive.

(Brendan at 00:39:29) Yeah. That's pretty inspirational to make it through that. I can't imagine it.

(Joel Beasley at 00:39:35) Well, I mean, it's just like you can either—it sounds cool, but when you're sitting there in the wheelchair, you just don't wanna be there. You're like, what do I have to do? Like, what do I have to do? How do I get up and not be in this thing? Like, where's the path? Because you get down about it a little bit. I actually ended up getting really overweight. I blew up to 300 pounds. So I was in the wheelchair, broken legs, got depressed, got really, really overweight. And then I'm in shape now. You know? But it was this whole mess of a decade for me. And then I'm like, every single day, I'm just like, at least it's not—at least I'm not 300 pounds in a wheelchair, can't walk anymore.

(Brendan at 00:40:19) Yeah. That, yeah. It sounds like quite something to have for a while.

(Joel Beasley at 00:40:25) It's like my flight's late, but I can walk. Yes. I win.

(Brendan at 00:40:31) Yeah. That's, back—you know, it's funny. I think it's that thing that lets you go that extra, extra mile. I have a similar—what you're saying about the flight's late. Once I—I woke up with the—I was in college. I left college for a semester to start a software company. And at that time, Y Combinator had just launched, and I was invited to a private reception in Cambridge. And I missed the flight because I had the flu, and I literally made it out later that day to make it to the reception that night because I was so determined to get there. And I think it—you know, like—

(Joel Beasley at 00:41:13) Made it happen, man.

(Brendan at 00:41:14) Yeah. If I hadn't gone through all the ordeals leading up to that, it was how hard it was to start a company and kind of learn entrepreneurship from scratch. I don't think I ever would've kind of used that experience to make it despite probably getting half the people on the plane sick.

(Joel Beasley at 00:41:38) You're like a little terrorist, Brendan.

(Brendan at 00:41:40) Yeah. Now that I'm thinking about it, I'm like, oh, what was I doing to other people?

(Joel Beasley at 00:41:46) The important thing is you made your Y Combinator. Oh, my goodness. Oh, man. We'll edit that out. Yeah.

(Joel Beasley at 00:41:55) I just got an email from your PR person. They say, edit, edit, edit. Yeah. Oh, yeah.

(Brendan at 00:42:02) It's funny because I got to meet Steve Wozniak at that event. So it was, you know, every moment was worth it.

(Joel Beasley at 00:42:08) And he remembers you as the jerk that got him sick. Alrighty. No. I'm just kidding. Oh, Brendan. So it was cool meeting Steve Wozniak? The Woz?

(Brendan at 00:42:18) It was. I don't—it's funny. That night is a total blur. I got into a long discussion with somebody, I think one of the founders of AbiWord, the Linux word competitor. But I sort of—that's what I remember from the night because I had a fever after that. And—

(Joel Beasley at 00:42:42) You were tanked on NyQuil?

(Brendan at 00:42:45) Oh my god. Yeah. That night, I had to go outside the city to crash at a friend's place in his apartment near his campus. And I just was—I was out of it for the next few days.

(Joel Beasley at 00:43:01) You're stumbling around the city. Everyone thinks you're drunk, but you're just sick.

(Brendan at 00:43:05) Oh, yeah. Yeah. And there was no Uber back then.

(Joel Beasley at 00:43:09) Oh, no. All of a sudden, you see—I'm gonna see Brendan riding Spot by Boston Dynamics. I'm drinking—oh my god. There's the use for them, by the way. Little Ubers. They can just ride the dogs on, put little saddles on them.

(Brendan at 00:43:24) No. That's something I could get behind. They're almost big enough for that. I have a friend who has a robotics company here. They make this robot that carries cargo for you, rolls behind you, called Gita. And it's funny because it's big enough that I always wanna put a saddle on it and ride it.

(Joel Beasley at 00:43:45) Oh, wait. Wait. Wait. Who's this guy? Who's this friend or girl?

(Brendan at 00:43:49) Her name is Sasha, and it's Piaggio Fast Forward that makes it.

(Joel Beasley at 00:43:56) Oh, that's amazing. I didn't even know about it.

(Brendan at 00:43:59) Yeah. It's—let me see if I can paste into the window a link to a video that she had. Let's see. Because the thing is—it, like, the demos they have are literally it carrying your shopping with you. It has this auto-follow technology. You put a little thing in your pocket, and it follows you around with obstacle avoidance.

(Joel Beasley at 00:44:23) Oh my goodness. I'm in love.

(Brendan at 00:44:25) Let's see. Media. Of course, it's an actual media page. This is the actual website, though.

(Joel Beasley at 00:44:34) Gita. Oh, man. Introducing Gita. Oh, wow. Hold on a second. A lot of up-close shots. Give me the—I know you got it.

(Brendan at 00:44:44) Picture.

(Joel Beasley at 00:44:45) Oh, I'll scroll down.

(Brendan at 00:44:46) Yeah. The problem is a lot of the videos on LinkedIn, I don't actually know how to share them out of LinkedIn.

(Joel Beasley at 00:44:52) You can't really at all.

(Brendan at 00:44:53) Yeah. And so she had a video she posted. There's this very kind of well-known bartender in Boston that a lot of the startup community knows well, and he used one of these to propose to his girlfriend. And so they had a video of their first, you know, proposal.

(Joel Beasley at 00:45:15) What a—I love this guy. Oh, man. The lights on it.

(Brendan at 00:45:18) Yeah. It is honestly—it's a—you know, the company is Italian. They make the scooters. And this is their robotics division that they put in Boston. And so it's got Italian styling, and I think that kinda makes all the difference in the world.

(Joel Beasley at 00:45:32) Yeah. I'm looking at it right now. It, you know what? It does feel very Italian style. It's that plastic stuff, that hard—the coating of it reminds me of the little Italian cars. Right?

(Brendan at 00:45:43) Yeah. Exactly.

(Joel Beasley at 00:45:44) Oh, that thing is so smart. Can it do stairs, though?

(Brendan at 00:45:47) I don't think so.

(Joel Beasley at 00:45:48) No? It's okay. Version two, it's coming. It's coming. Put it in the pipeline. Yeah.

(Brendan at 00:45:52) I mean, you need bikes for a stair pretty much.

(Joel Beasley at 00:45:55) We have a—we have a pivotal story for it or an epic. We have an epic. Just the stairs.

(Brendan at 00:46:02) I remember they brought one up on stage at the same—the robotics conference I was speaking at, and they definitely used a ramp on the stairs. So—

(Joel Beasley at 00:46:13) Oh, no. What conference was that?

(Brendan at 00:46:14) Limitations. That was the Robo Madness, put on by Xconomy.

(Joel Beasley at 00:46:22) What? That sounds amazing. Robo Madness?

(Brendan at 00:46:26) Yeah. Every year, well, you know, iRobot is nearby, and Boston Dynamics is nearby, and you've got all this. So there's so many robotics companies around here. They put this on, and they basically get a bunch of people talking about AI and robotics. And last year, Stephen Wolfram talked, and the founder of Boston Dynamics talked, and the CTO of iRobot. So it was—it's a pretty good conference, and they held it last year at least at Google. So there were great facilities for it.

(Joel Beasley at 00:47:02) I gotta be calling these people—and that was real proper English. I need to call these individuals. I'm sorry. I got excited for some reason. I need to call these people and say, I need your iRobots. And what does iRobot produce that's really ridiculous and cool?

(Brendan at 00:47:23) I mean, I think it's been normalized. Right? They mainly produce the Roomba and all of its ilk. But their military robotics division split off from the company thanks to an activist investor. So now there's another iRobot that just makes all the things like Mars rovers and bomb diffuser robots.

(Joel Beasley at 00:47:51) Yeah. I mean, it's really difficult from a PR standpoint when we're gonna clean your floors and kill people.

(Brendan at 00:47:58) It's like—

(Joel Beasley at 00:47:59) Yeah. It's tough.

(Brendan at 00:48:02) Yeah. I had a chance at a dinner to sit next to the inventor of the Roomba, and he's a very down-to-earth kind of Woz-like guy. And he was telling me about his new project, and he's launched it now. I forgot what it's called, but he wanted to make a robot to automatically do your gardening for you.

(Joel Beasley at 00:48:23) Oh, yeah.

(Brendan at 00:48:24) And he's launched it now. So there's another company now, another iRobot for gardening.

(Joel Beasley at 00:48:30) Oh, of course. I want—if I sit next to him, I'd be like, hey. So, how'd this come about? He'd be like, it's the only chore my wife gave me, so I automated it. Oh, man. Wow. Brendan, I love talking with you, dude.

(Brendan at 00:48:47) Yeah. It was fun.

(Joel Beasley at 00:48:48) Yeah. This is amazing. It's real simple. Thank you so much for listening to the Modern CTO podcast. Share this. Get the word out. Thank you guys so much. I couldn't do it without you. I appreciate it. You guys are the absolute best.