Cole Dermott - Co-Founder @ Locus (YC F25)

Welcome to another YC Wisdom pod. We have Cole Dermott with us. He's the co-founder at Locus. It's a YCF '25 batch. He's a previous engineer at Coinbase. A fellow Canadian, right, Cole?

Yep, yep. Born and raised in Ontario.

Ontario, nice. Shout out to Waterloo. I heard on the podcast not so long ago that SF was a Waterloo rapper. What do you have to say about that?

I'd agree. I've not agreed, but we got a lot of people down there. It's kind of crazy. You'll be down there talking to interns, and it's like every 5th to 6th person you meet is like a Waterloo student.

That's crazy. Why do you think that is? Why? Why Waterloo is what it is?

It's their co-op program. I don't really have another question in my mind. Every single student from Waterloo ends up by the time they're graduating, working between 4 to 6 full-time jobs. Jobs in the industry, right? And that's just a huge structural advantage if you're walking out of doing, you know, a degree for 4 years, right? You have a year and a half to 2 years of practical work experience versus everyone else having zero in a market that's defined by essentially what you know. That's more valuable than anything.

And who are you, Cole? What shaped your brain?

A lot of things. That's a very broad question. Uh, I'd say I'm very, um, Let's back up a bit. So since I was a kid, I've always been really motivated by financial systems, right? Um, so growing up, I was super passionate about economics, investing, and everything like that. When I got older, um, I started to get exposed a little bit to the entire cryptocurrency world. Um, saw one of my friends back in high school make an absurd amount of money off Bitcoin just because he happened to set up a Bitcoin mining rig with a bunch of old Windows phones. Um, and got pretty interested in the space, started investing, whatever, Flash forward, I actually didn't code at all before university. I was someone who's going to go into business. I've always been fairly entrepreneurial, ended up getting into Waterloo CS, and so very quickly started to learn the technical side of how do I actually implement the ideas I have. From there, was in the program for a while doing my thing in my degree and ended up joining up with Waterloo Blockchain, which is a large builder syndicate at the University of Waterloo. It's the largest nonprofit student blockchain org in North America. So I joined up with them, went to a bunch of hackathons with them, started to win, realized pretty good at the developer side. Then got a job at Coinbase working on their Coinbase Business platform, doing stablecoin engineering, stuff like that. Saw kind of the gaps firsthand throughout that time in agentic payments right now and started to realize that I've had this idea burning in the back of my mind. I've always wanted to start something. I don't think I do well working for other people. I'm a very independent person. I come from a very small family, so I had to be that way most of my life. And when I saw the opportunity in a space that I knew better than most people, uh, or anyone really, uh, I took the leap and started Locus.

And you say small family. Who are your parents, uh, by name?

Or like, um, yeah, I don't care about their names.

Yeah, were they entrepreneurs? Were they, um, what was their profession, right? So I always like to explore the nurture vs. nature question. Uh, so what genes they gave you, and also how did they brought you up? My dad was a an entrepreneur. My mom taught me how to be obsessed with books, right? She put, uh, Think and Grow Rich in my hands at like 19. Uh, just like I was a busboy and a server back in these days in Canada, I was very entrepreneurial as well, already launching businesses. And, uh, that's what shaped me. What about you?

Pretty similar. Um, so both my parents actually worked in sales their entire lives. Um, so a variety of things. Um, and so I learned a lot seeing that firsthand. It's It's kind of interesting, my dad's actually a university dropout and ended up climbing up the corporate ladder in the sales world. So I saw a lot just like firsthand of how important interacting with people is. Uh, separately from that, they both honestly encouraged from a really young age, uh, the entrepreneurial side of me. So I can remember back in— I was a huge Nintendo fan growing up. I really liked, uh, like old Nintendo consoles. It was just something I was passionate about.

Super Smash.

Yeah, yeah. So I remember back Exactly. I remember back in the day when, um, one of their consoles come out, they came out with these like figurines called amiibo. So I was around like 10 at the time, um, and I ended up setting up like a full eBay scalping business basically of, you know, reselling these figurines at the time and earning like very small profits. But, you know, my mom would, uh, encourage me constantly. I still have memories of her taking me, you know, immediately following school to whatever store just to find the new drop before it sold out so I could resell it on eBay. Effectively like chauffeuring me in that. So they've always been pushing me towards more of the entrepreneurial side, I think. And then separately from that, throughout, you know, most of my life, my dad was someone who had a lot of ideas and I think was always chained down by the corporate world. And so firsthand saw, you know, pursuing this, especially at the age I am now, is the pragmatic thing to do, not the hopeful one, right?

And your stint at Coinbase, did it change your mentality? I think that was a big win before Locus.

I'd say so. Um, not so much changed my mentality, it more just was firsthand proof of the things I already believed up to that point. Um, at Coinbase, I was able to like make myself very, very known for, for an internship, right? Um, I was able to like, at the end of the internship, the CEO retweeted, uh, my end of internship announcement. Coinbase's official Twitter account follows me now. I was able to do some very, like, high-level, uh, strategic work at the company for someone who was, like, just starting there, of a, you know, small intern pool. And so in general, it really just— I've always had the belief that if you're high agency, you put yourself out for— out there, and you really just do things, as dumb as that sounds, you will go places. Because the average person doesn't The heuristic level of like, oh, I need to go talk to someone or do something like that level of intimidation stops most people. Right. And so if you're someone who just does that, you go way further than most people.

Especially when you're against polished kids that are Ivy League and all of that.

Yeah, exactly. Like, you know, I'm, I, I think I'm a little more unstructured, let's say, than some people, probably not as organized as some people, but I'm also one of the hardest workers out there who's louder than most people. And doing that is. I really think how you win in the end.

And looking at your YC batchmates, right, because YC sort of stopped your stint at Coinbase, or else you would probably still be there, uh, what do you see in yourself that your batchmates don't have?

It's an interesting question. I don't know if relative is the best way to ask that. I think what I'm very exceptional at in a vacuum is, um, Look, I think I'm a very, very strong communicator. That's something I've had my whole life. I've been very good at conveying my thoughts and feelings into words to people and making it in a very concrete way that they can either execute on or to drive emotion from an audience. Separately from any of that, I'd say I'm very pragmatic. I tend to be very, very logical in the way I approach things. Most of my decision-making rests on data above anything else. And I think I do a good job at removing myself emotionally from most situations. I'm acting in. Uh, past any of that, you know, I, I'd say third fold is that I'm very good at seeing the big picture. Um, I oftentimes, and this is like something I'll fault myself on firsthand, miss the little details when things are moving around. But in terms of seeing the big picture about where things are heading, where we need to be, where I should be sitting, and where, you know, in general the market's heading, um, I think that's something I have a pulse on better than, you most people I've spoken to.

And let's ask the other end of that question. Where do you see them more talented than you are? Because it's extremely talented people in YC, right? So what do you see in them that you're like, holy shit, I need to take notes here?

So something I admire about a lot of these people that I can't do is I'm an incredibly routine-based person. If I don't get my 8 hours of sleep and work for a consistent schedule, my entire work life habit, like ethic just falls apart completely within like 2 days. And so one thing is I just see some of these YC people and they can work more than I can. My body just can't do that. And that's, I'm talking about the people who will legitimately work like all night, several nights, stuff like that. I've just never been a person capable of that. I fall apart. Second half of things is I think a lot of them are more organized than me. That's something I've struggled with to this date is in general, Locus is a very unstructured company. I'll be the first to admit that, you know, we, One of my biggest hiring philosophies is I hire operators and owners, and a large part of that is because the work is so unstructured here in terms of things need to get done. There's so much work to do. You just need to go and oftentimes find things to do yourself. That's on me for not being a better manager a lot of the time. Right. And so those two especially stand out as places where I feel those gaps in my day-to-day operations, and I'm actively trying to work to fill them through throughout my time here.

I want to double-click on the 8 hours of sleep. Jeff Bezos slept 8 hours, and I don't think that Elon playbook is sustainable. I think Elon has the genes plus the multiple years of adaptation to run on that. So a counter-question, right, from a first principle and scientific perspective: do their brains run at times too? Like, I, I sleep like 8 hours, 9 hours of sleep. My brains run at least 2 times faster than most humans. I have 15+ meetings on a daily basis, so I need like some time to sleep and digest all of that, right? Like, so do you feel you're comparing apples with apples here?

Maybe not. I mean, look, it's pretty similar to the way you look at it. It's the marginal benefit of the additional 2 to 3 hours of work I could get out of doing in a day where I push that hours of sleep down to like 6 or 5 to get those additional hours. It causes so much detriment to the, you know, 16 hours of work that I'm doing the next day to the point where it's just not worth it. And like, I don't know if it has to do with how my brain works necessarily fast or anything else, right? It's just—

It's diminishing returns.

Is optimal. Yeah, it's exactly. My entire goal is optimize every waking hour I have. And if 16 hours of work, I can get accomplished what I feel like is closer to 30 hours of work with, you know, leveraging AI tools, the way my brain moves, everything like that. Versus if I work for 18 and the next day I'm only getting across, you know, 20 or 24. That's not, uh, it's not a fair trade-off at that point. It's not worth it.

Yes, there's a counter-argument, like depending on if you're an out-night owl or something like that, that working at night is more peaceful, right? There's less noise around you, less requests, less DMs. Um, but yeah, do you feel some— sometimes you feel dumb, like working on some stuff that you feel that an agent could do?

Occasionally, I guess. Honestly, I think I've managed to automate most things that an AI agent would be capable of doing for me by this point.

Do you feel agents will fully automate you someday soon?

No, because I think taste and creativity, as well as taste and creativity, in the end will be tantamount. AIs are numerous but identical. Individual human taste is something built by lived experience that an AI can't replicate.

But we can feed all these lived experience to AI. So what, what is stopping an AI to be creative and, uh, tasteful?

Can I feed all my lived experience to an AI? Like, there's a lot of ages that I wouldn't be able to describe by this point, a lot of experiences that more so sit implicitly into my subconscious rather than explicitly outward-facing, right? Um, I don't necessarily believe that, as of now at least, uh, we can train an AI— like, we can nurture an AI to be the exact same as us. Um, In the end, it's still— you're fine-tuning it, you get it closer, right? And it gets better. But there's still some element of you need a controller at the end of the day to sit there. That's my honest belief. I think we're going to get to a point eventually, maybe, where they're good enough to, say, run their own business and they can do that profitably. That's pretty core to a lot of the work we're actually doing now at Locus. It's just you still, in the end, need that element of human taste to sit at the— you imagine a knot. The thing binding that knot together is the human taste at the end of it.

Yeah, sure. And not sure if I get the knot analogy. The problem with subconscious is I feel it's a word to describe what we don't fully understand yet. And I would describe the brain more as a database, although like a bunch of people would disagree. Like, what's your conscious take about you tapping into your subconscious? How can you scientifically quantify that?

That's my point though. There's so much data that you accumulate as a human that I agree with your take on database, right? That is effectively how our brains work. We store information, we access it randomly. Like, it is exactly like that. Um, the problem is that a lot of that info is deeply, deeply buried in your head, and there's no exact way to extract information out of someone's head in a way that can directly translate to, hey, I can give this entity everything that I have lived through, everything that I have experienced. And suddenly it can act exactly like me.

Right. Talking about Locus, right? Because I want to take a bit of time here, Billy. We have like 10 more. How big is the agent take payment market? I believe there was an announcement with Visa not so long ago. I believe there was 402X by Coinbase. I think they sort of failed to some degree. Can you give me like an overview of the market? Because obviously agent take payment is gigantic. So, but eventually now I think it's pretty weak. As a TAM?

Yeah, you kind of hit the nail on the head. It's, if we're looking out like 10 years from now, we're talking several trillion dollars, right? It's an incredibly massive market. You look at what you can automate, you can automate most of consumer shopping, you can automate accounts payable, insurance remittances, payouts. There's tons of examples anywhere where you have payments already happening today that aren't purely structured by if-then statements, but rather conditional decisions along the way that humans are making. That's where Agentik Payments can be leveraged and be valuable. Right now, though, to your exact point, yeah, the market's tiny. We're seeing like maybe $50K to $100K of volume a day, of which probably around $20,000 to $15,000 is fully legitimate in the sense of real people using this for something that is accruing value. So it's tiny at the moment for sure. It's also weird. It's a very fast-moving market where we're seeing a huge land grab for effectively very very little real estate that exists, right? Um, however, look, it's growing pretty rapidly too. Um, I think there is something here. It's just the pace of adoption, uh, is probably going to be a little slower than a lot of the current players are expecting.

How do you position yourself next to 402? And I believe Stripe has something as well, right?

Yeah, we build on top of them. Uh, so, you know, MPP, which is Stripe's thing you mentioned, we're like around 25 to 30% of the weekly volume on that platform from our endpoints. So at Locus, we didn't just build like the payment infrastructure, we also built things for agents to buy, right? Then separately, pretty soon we're releasing ways for agents to earn. So we kind of just think ourselves as powering the agent economic flywheel and sitting to enable the use cases that are actually feasible and realistic today where there is demand.

And why are you the right guy to build this? Because I'd invest in you guys. I'd love to hear about the last fundraising numbers, but I think I'd need to be very patient with my investment, right? Like Jeff Bezos style. So, um, yeah, like, tell me a bit more about that.

So I'll lead with I can't disclose numbers. I won't. You can ask me off the record after the pod if you're interested. Um, we haven't announced our fundraise publicly yet. Um, but past any of that, why me? So look, um, above anything, I think I'm someone who can see noise and see signal and actually direct myself towards the correct thing. In all of this. I led with I'm very good at seeing the big picture, right? I think the space has a lot of noise, to be totally honest. It's what happens when you have a frothy market with only $100,000 in daily volume, of which a very small portion is legitimate, right? And then you have around 40 players all trying to capture that small amount of real estate. That's going to create froth. You can't do anything about it, right? We have taken a very unique, distinct approach compared to the competition. I think what everyone has done is they've built a lot of ways for agents to pay, and they've gone and built a lot of supply. And the stance at this point is mostly let's wait for, with some exceptions, and I'll call out the ones as I go here, but they've kind of waited for demand to come. The approach we're taking is very, very different. The approach we're taking is we're incredibly focused on creating our own demand by giving use cases that are actually feasible and achievable right now for people to do. So like an example of that's Locus Founder. Locus Founder is a product we have coming out on Monday, June 15th. It's an AI agent that builds and runs internet businesses for you and can be entirely driven and ran over text. It has access to make payments, receive payments, send emails, run outbound campaigns, post and manage ads across Meta and other platforms, and much, much more, right? So you can think of this as the earning part of it. And so like the structure Locus went through and why I'm pretty confident we will win in the end is, look, we started off building the infrastructure. Demand was low, so we went and built supply. We built supply in a bunch of places and some of the most used supply. We saw very quickly, we accumulated a lot of data from doing that. And now it's at a point where it's fairly low touch for Locus, the organization, right? We just have this data coming in every day. Now using that data and using the payment infrastructure we built that we're confident is best in class, but you know, there's just not enough demand right now to get sustained usage. We went and built this consumer application on top of all of it that will act as the demand driver and generator. And so, you know, Net-net, why we win, it's the insight that if the market's not ready, we can't sit on our laurels. The way we win this is we make the market ourselves and we bring the future to now.

I love it, dude. So you said fully automated agentic outbound campaign on the 25th, right?

15th. So this is on Monday.

Oh dude, I'm gonna be one of your first users. So let's go for that. Are you a solo founder? So I know where that leads, but go ahead.

Yeah, I feel like this is— what can I say about it? I used to have a co-founder. We kind of disagreed about the direction of the company and direction of the space, debated a lot about it, and thought the best step forward was for him to go and explore his own new things and for me to stay behind at Locus. Incredibly smart guy. I'm sure he's up to really big things right now, and his next company will be massive. His name's Elliot Lee, if anyone wants to look him up. But yeah, now at this point, I am effectively, I guess, operating like a solo founder at Locus. But also, look, I hire operators and owners. I like to think everyone on our team is a founder of some small part of Locus.

And as an investor, I mean, it's not— it could be a yellow light, right? What could— like, how can you outsource the decisions to like more players so that there's a sounder chance that it's a good decision, right? Like, I'd prefer to have 2 or 3 YC brands rather than one.

I'm a little— can you elaborate on that question? I'm sorry.

Yeah, when you invest in a founder team, right, you know, one of the principles of YC is to have multiple founders, at least one co-founder, because it's a sounding board, right? It's like, it just makes more sense, and there's higher statistical probability that you guys make decisions if you're two. Um, do you feel you would have a mechanism for that even if you don't have a co-founder?

Yeah, I mean, that's why, uh, my hiring process— I've really focused on this a lot. Like, I'm very confident in the team we have here, right? I think generally speaking, I would push back a little bit on some decision-making heuristics here, because if you are confident enough in your decision-making and you're strong enough at doing it and consistently based off heuristic data and insights you've accumulated, you make the right decisions. It can be worse to have multiple cooks in the kitchen. You can have people who have completely polarizing viewpoints from you and lead you astray from decisions that consistently would have proven themselves to be correct. Regardless, as of right now, um, I have team members who I put a lot of stock in, right? Like Locus founder, the idea behind this product we're releasing on Monday, it was, uh, you know, our founding engineer's idea, right? He's someone I respect a lot. I've worked with him a lot in the past. Uh, when I do have doubts about my decision making, I open it up to the team. I tell them about, hey, here's my mental model, here's what I think we should do, here's why. Where would you push back? Where do you think I'm wrong or what do you think I'm not thinking about? Right? I don't know. That's my personal stance on it.

Right. Your own GTM, what does it look like?

Essentially creating for Logos founder at least as wide a top of funnel as possible. So I think it looks a lot like looking at this. So first off, our market, for a Locus founder, right? We're talking about either young hustlers. You can think the people used to do sneaker bots, sell Pokémon cards, anything like that. They're probably what, like 15 to 25, maybe 30? Not super technical, super motivated. Yeah. On the other end of that spectrum, we have the generally boomers. They're older. They've probably worked a corporate job for a long time. They've always had an idea sitting in the back of their mind, lacked the resources to pursue it, but they're really interested in AI and they want to kind of be part of this revolution they see happening from Yeah. So what it's looked like for us is looking a lot at A, where those people live, right? So if you think of those two very broad markets, you can identify the young people as generally living on like short-form content areas. To be honest, that's their main media they consume. And if you look at the higher end of the spectrum, the boomers, it's generally going to be Facebook and Meta as well as sometimes LinkedIn, right? And so what it's been is casting as wide a funnel on essentially each of those areas is what, or what it's what we will be doing come Monday, uh, to clarify.

Any cold outbound?

Uh, cold outbound for a product like this for consumer doesn't work super well. I mean, there's been some, mostly to— you think of a single person as a distribution channel where they have like an audience.

Why go consumer? Why not B2B? Isn't that way more sexier?

Um, look, we have B2B operations, uh, but also right now in Agentic Payments, it's It's so early that demand has not caught up yet. And again, I think if you look at any sort of payment company long term, adoption didn't really come from big businesses starting to use this tech. Adoption usually came from a guy in his dorm room who was doing something stupid online and then wanted to start using this, right? You kind of have to start grassroots to get adoption in payments from everything I've seen. Look at crypto in the past. Look at Stripe in the past, how they started everything. Um, I don't know.

One last question for you, dude. Um, wouldn't the hockey stick moment in your vertical be like Anthropic launches its own agents very successfully with a bunch of integration and/or OpenAI and/or Gemini, and I would be the startup to partner with them to give them the agentic infra payment rails?

Yeah. Um, look, like, I think that's something we want to do in the long run. That is like the North Star goal here. Uh, but also talking firsthand to a lot of labs. I just don't think there's interest right now in them adopting it just yet. They're often focused on model development and less so innovations in terms of like financial interaction, right? Right. We're still dealing with this mental model of can an agent do financial analysis? Like, that's only recently something we've passed, right? Now, an agent automating all your outbound payments, it's not where these people's heads are at just yet. Uh, we got a little bit more time before we're there.

Yeah, and you need to find a business model that will bring short-term revenue to motivate the investor and invest in you. So I agree with that, uh, model very Cole, where can people find out more about you?

Uh, you can follow me on Twitter, x.com/coldermo, C-O-L-E-D-E-R-M-O. And then separately, uh, locusfoundry.com. I honestly— learn more about me if you want. I think the most interesting thing you can do is go to locusfoundry.com, L-O-C-U-S-F-O-U-N-D-E-R dot, uh, dot C-O-M, uh, and try the products. I, I think you'll, you know, experience something pretty cool and pretty different than you've ever seen before.

That was Cole and Charles for another