Jake Aaron Villarreal: I'm Jake Aaron Villarreal, born and raised in Silicon Valley. I'm here to take you behind the scenes to share what it's like to be a startup founder. The journey they're on, the problems they face, the products they build in an effort to make our lives better. I'm excited to have with us today Tom Firth, co-founder of Cotera. Tom, welcome to the show.
Tom Firth: Hey Jake, thanks for having me on. Appreciate the time.
Jake Aaron Villarreal: Well, thanks for joining us. A little bit more about Tom. He is the co-CEO of the startup which is a Y Combinator-backed organization. They're building AI agents that help businesses automate work at scale. So Tom, I know when we had our pod prep call, we talked a little bit about you and more in depth. You came from England, Oxford. Whereabouts are you originally from?
Tom Firth: I'm actually so I'm actually from Kent which is just south of London but I was living in Oxford for about six years before I moved to the US.
Jake Aaron Villarreal: Very cool. [clears throat] So yeah. So you got a different perspective from not being born and raised in Silicon Valley which is where I'm from but [laughter] I'm sure you've got a lot of technical knowhow and your experience so far of what you've started and where you're at today. So we'll dive into that. Before we do, give us a little background of yourself. What kind of shaped you into becoming an entrepreneur and more importantly how did you get into tech?
Tom Firth: Yeah, it's interesting as as you kind of called out I, you know, I was not born in Silicon Valley. In fact, like there's not really much from my childhood that would have suggested I was going to end up working in startups. I think like nobody was an entrepreneur in my family. One thing I always was was just sort of like super curious and pathologically independent is the way I like to put it. My mom always has all these funny stories of me as a kid, just refusing to do things the way that anyone else did them. And I think between that, some sort of like personality traits, and then gradually getting exposed to kind of like the software world as I as I grew up, yeah, it took took me that way. It wasn't until I was in my early 20s, I think, that I knew for sure I wanted to start like a tech company. So, it took a while to find my way there, but yeah, once I did, no, no going back really. It's just too much fun.
Jake Aaron Villarreal: Yeah. Well, you started and this is... is this your first startup?
Tom Firth: It is. Yeah. First time founder. Although three years in now, so starting to starting to feel, you know, got a few gray hairs going. [laughter]
Jake Aaron Villarreal: Well, for the listeners, you couldn't tell. He's got a nice set of hair and none of it's gray. So when you began the journey of building a company, building a startup, what was it you identified in the market that you and your co-founder saw that you felt like, "You know what, this is a good time to build something" and to really apply to Y Combinator and and get in at a time where there wasn't a lot of in-person activity going on.
Tom Firth: Yeah, it was one of those COVID batches. It was the winter 2022 batch and I think it was the largest batch YC ever did and also like one of the very few fully remote ones, and yeah, it was an interesting time to start a company. But so both me and my co-founder worked in data before this. I used to run the data team for a SaaS business that built like an equity research product in the UK, and Ibe, my co-founder, was the first data science hire at Peloton. And both of us had really spent the previous you know, all of our 20s basically working on helping companies use data, whether it was for product or for decision-making or for automation internally that had sort of been like the bread and butter of what we've done in our careers.
And we were one of these startups that I think didn't have like... some startups go in and they've got an absolutely crystal clear idea of the thing they're going to build, and I don't know if we were one of those at the start. But we definitely knew that we were both just data nuts basically. We loved writing software, we loved working on data, we really believed in sort of the power of data done well to help companies like run better, and we knew we wanted to work on something doing that. And what we got into YC with was a product that we don't work on anymore actually, but it was still rooted in the kind of like yeah, this kind of core belief that companies that do data well have an advantage, and very few do do it well, and we want to help them do it well.
And it did kind of like twist and turn a bit as things went on. But the biggest twist for sure was as all the AI stuff took off. And we had already built this kind of platform, a foundation layer that helped us run more traditional machine learning models on top of people's data infrastructure to help them automate stuff. And the AI boom kind of slotted really nicely into that. And very, very quickly, our customers, you know, were asking us to do the same things we were doing for them already, but with LLMs layered on top. And so that kind of worked out really nicely. And that's sort of what the business model is today. We basically integrate with all your data, all the sort of systems that you want to take action in. And then we have this kind of decision-making engine, you could call it, built with LLMs on top of that to to automate work.
Jake Aaron Villarreal: Yeah, I like that how you titled yourself as the "data nut." That's that's one I [laughter] got, it's one I'm gonna start using more often with our own team. "Get the data nuts. Come over here, guys. Let's get rolling." But yeah, no I it's, it seems like the timing was right for your backgrounds. I mean, it's all about data with AI.
Tom Firth: It is, isn't it? Like there's kind of... I think of the AI space as... I mean I'm not the only one to think this, but there's a few sort of clear tiers I guess you could say to the ecosystem. And the model companies and the model vendors is one thing and obviously that data is very important there, but for for a different reason. That's really about training data. And where we work in terms of applying these models to business, it's all about context. And like the models are pretty smart, but only if you can get the right context and the right information to make a decision or or something like that. And it's all just data pipelines. It's kind of all the same stuff. It's not really that different to building a BI dashboard in some ways.
Jake Aaron Villarreal: So yeah, talk a little about context because you said that a few times in our first call and you know we look at AI and we talk about models and we talk about infrastructure and different companies that are building niche products in in different business lines. Context is a big part of how you get the benefit out of AI. What does that mean to you?
Tom Firth: Yeah, it's one of these topics that's it's very simple really and that makes it surprisingly slippery, but it's literally just... I, the best intuition I can give somebody for it is if you, if you, if you hired somebody very smart into your company and you asked them to do some task that was very specific to your company, like I don't know, like "go and refund a customer," really, really simple example. It's actually pretty hard to know how to do that in a given company even for somebody who's smart because what billing system do I use? Like what are the rules for when I can refund them? Like which three systems do I have to update if any after I've done the action? And all of that information, this kind of like specific information about the circumstances you're doing an action in. That's kind of the context in this case. And without that context, doesn't really matter how smart you are. You kind of just don't know where to begin. And I think that's exactly the problem with LLMs. In terms of like raw horsepower. They're actually pretty smart. They can, if if you set them up right, they can answer a lot of pretty tough questions, but but if you don't give them the right background information to to do it just right for your specific circumstance, then then they're kind of like useless basically. So, I would say the like yeah, that's a very practical definition of context, but a very, a one-liner is just background information relevant to the task, I guess.
Jake Aaron Villarreal: Yeah, got it. Yeah. You talk about prompt engineering for like just the layman who's going into ChatGPT or yeah, you know, any other LLM and it's, you know, give it some context about what you want it to do and what you're asking it to do and what role it's in to be able to deliver the content that you want. And even at the basic level, people don't really understand that quite yet. But on the business side, yeah, on the business side, it's there's much more complexity because you're talking about, you know, how a company operates and the workflows of an organization or what's the standard operating procedures for companies and then taking technology and automating that and doing it in a way that you can't disregard the complexity of what you're trying to solve. At the same time, it's all about the context of what you're wanting it to do.
Tom Firth: Yeah, it's exactly right. And it's both very simple. Like it's quite straightforward. It's like, "Oh, if we want to automate this task, we just have to write down all of the details that are required to like, you know, required to do it. Like what would I have to write down for a new intern to do this successfully as well." So it's very simple in that sense, but at the same time it's actually very difficult because it's amazing how much stuff actually just happens through sort of tribal knowledge. And you know, people who've just been there for 5 years kind of like pick things up and it's not written down anywhere and sometimes it's actually pretty inconsistent too. It turns out that the refund policy we thought we had, oh everyone everyone actually has a different thing in mind and five people are doing it five different ways and no one can agree on like the right way to do it. So like business is messy. It's like the real world in general is messy, but I think that's a big challenge with these things, trying to trying to make it less messy.
Jake Aaron Villarreal: Yeah, it's interesting you bring up tribal because that's something that we we've recently heard a lot more about is like what's the knowledge that people internally understand in the culture in a company of how things operate and then what's, what is the language they're using? What do they believe it is? And if you're talking about automating any process in a company and using AI agents, like you know there's certain scenarios where it's pretty straightforward and there's a lot of edge case edge cases where it's not that straightforward at all and you have to really dive into the detail to make sure that you're understanding the outcomes of what a company wants AI to do for them and then be able to understand that, document it, and create it and so you can deliver based on what their outcomes are and speed to deliver and you know making sure they're paying for what they want. And and you what what I think is really interesting about your company is you kind of were contrarian, you took a different approach where a lot of startups went out, they got funding and they said "okay we got the money let's go build, let's go hire, let's go scale and we'll figure out the market as we go." But you took a different approach. Your approach was "let's keep it lean and like legitimately lean," like you're at what eight employees today.
Tom Firth: Eight employees at the moment. Yeah, exactly.
Jake Aaron Villarreal: Okay. And what give us the insight because you didn't really have like a map to say that was the right choice to take? You just you innately made that choice. What was it that you felt that you wanted to take that approach?
Tom Firth: Yeah, it's a very it's a very interesting topic. I think I think I think there are many successful ways to build a company. And to a large degree, I think one of one of my pet theories that I've evolved is that the kind of like there's this sort of founder-strategy match that's required where different founders have different personality types and different personalities like suit certain like methods of execution well. And so you can, there are definitely people who have gone out and just raised loads of money on like just like hot air basically and then been like "yeah let's try and figure it out afterwards." And that's something I find really hard to do. For for me, for me to want to like, for me to sell something, I really have to believe it myself. And for me to really believe it myself, I've got to like fundamentally believe that the business is going to work basically.
And the thing that I, like I don't think you need a lot of money to figure that out. I think you basically need to go and get customers and listen to them and serve them well and build a product that they really love. And I think you can do that without without raising a load of money. And we raised, I mean it's not like we didn't raise, raised quite a bit in my in my books, just under $7 million in total. But yeah.
Jake Aaron Villarreal: Yeah.
Tom Firth: So, I don't know. I I like that, that kind of always made the most sense to me. It's kind of like get the capital you need to like answer the really hard meaningful questions that you're going to have to figure out. And I think it's easier to figure those questions out early than it is to kind of grow a big organization then try and retrofit them afterwards. And then, yeah, but things are starting to look very good. So, we are eyeing up a a Series A. I think mid 2026 or something like that is probably on the on the roadmap.
Jake Aaron Villarreal: I say nice. Well, [clears throat] talk to us a little bit about the problem you're solving because there's a lot of different approaches when you talk to companies about what you can do with AI and how you can help support their the outcomes they want. Dive into that a little bit. Yeah. How are you different?
Tom Firth: Yeah. I think the key thing that differentiates Cotera from most players in the market is that we're taking a sort of like fundamentally horizontal approach to how we build, help people build agents. We're really focused on this kind of like data infrastructure and tooling layer where we think that the thing that... the real problem that most businesses need to solve to to to use agents effectively is basically all around the data and the tooling. So if you have access to all the data i.e. the context, and if you have access to the actions you you you need to take within your kind of software stack APIs basically, then getting one of these off-the-shelf models to do the right thing is actually not that hard. And the model vendors are building the models. That's great. That's not really us. And we're focusing on the other half of the equation, which is basically like, can you get your data infrastructure and tooling to a place where where the models can kind of speak for themselves.
And yeah, so we've taken that like very sort of horizontal approach. It's a slower approach in some ways because, you know, I think the easiest thing from a go-to-market perspective in the early days would have been to sort of double down on like, "Yeah, we build this agent that does this one task and we're going to sell that one agent like a sales agent or something." I think we just believe that's not how the market's going to turn out. I think all of those kinds of products are going to turn out to be really, really easy and quick to build on top of a more horizontal platform. And so we kind of took this approach of building a more horizontal thing.
Jake Aaron Villarreal: Yeah. Yeah. Wow. Yeah. There's multiple different views on that, but you're finding success in your approach. Give me an example. I'm a customer out there. I've got loads of data. I've talked to a couple different vendors, maybe some big ones like Accenture and others that are smaller. But, you know, we want to automate parts of our business. We want to do it with AI and we want to do it in a short timeline where it's not going to take forever to see the result. What do you do? What's your approach to to a to coming into a company like that?
Tom Firth: Yeah. Well, one of them is quite funny. If you're talking to Accenture, then the timeline thing is guaranteed to be slower than us because they're they're not very [laughter] quick. They have got more bodies than us, I suppose, but but they're definitely not fast. It's it's it's really simple. I think the the consulting companies like Accenture and stuff I think tend to sell these things as these like transformation projects where it's always this sort of like large scale like "how like what's the business going to look like before and after AI?" And I and I think in reality the way these things actually get stuck into companies is you find some project where it's some combination of valuable enough and kind of like painful enough that some small little team, some sort of niche within the company, is ready to puts some money down, some of their budget down just to like solve the problem. And you start with one really, really concrete use case. You make that thing work end-to-end and then you use it as the kind of as the pitch to the next team basically, like to expand within the company.
I can give you some really concrete examples. Like Bilt Rewards, the credit card company, is one of our earliest customers now. Like we we love working with those guys. And there's probably double digit like workflows that we automate for them internally. And one like really stands out as quite a fun one is, I don't know if you've used the product, but basically you can earn rent on, you can earn points on your rent payments through it. And if you, if you're one of an out-of-network cardholder, sorry, your your your property is out-of-network, then then the way you get points is like a little bit more complex. You have to basically send in like proof of your rent payments to it. And for [snorts] some, some people make their payments through Venmo, which wasn't officially supported, but they would kind of like honor it in support. And so every month Bilt would get like thousands of these tickets being like, "Here's a picture of my Venmo receipt. Please can you like give me the points?" And then somebody in support has to go and like look up the points in another system and confirm they haven't already been issued and confirm it's like a real rent payment. And there's five or six of these steps. Then you kind of sign off on the points.
Turns out you could build an AI agent to do that very effectively. Everything from kind of reading the message to extract information from the the Venmo receipt to looking it up in the extra systems to reconciling it with previous payments. And you know, we turned something that took like it, it was not far off a couple hundred hours a month that they would spend of of people kind of like going through these down to a basically a 20-minute once-a-week review where they get like a big list of like "here's the ones we're going to approve," hit okay, and off you go. And uh, so it's things like that, like find some, it's a very, very specific pain point that only Bilt have that problem. But it really matters to them, and that their team was like sweating over that particular problem because it was very valuable to them and very frustrating. And if you can find the stuff like that that people have got, then actually yeah, even fairly big companies can move pretty quick to like solve those things. And and then you get a lot, if you solve it successfully, then you get a lot of goodwill to turn to the next team and be like "hey we'll solve your problems too" and that's what we did.
Jake Aaron Villarreal: Yeah, really cool. You know, I think for a lot of companies, it's about getting your foot in the door, too.
Tom Firth: Oh, yeah.
Jake Aaron Villarreal: You know, the one approach is, you know, marketing, founder, email, you know, events, like what's worked for you? What's worked for you as a company that's getting into big organizations and having success?
Tom Firth: Yeah. Yeah. It's a great question cuz we're we're sort of like halfway through evolving a little bit here as well. Where up until about 6 months ago, I would say the only thing pretty much we did was was basically cold email and and we we rocked up to a few conferences and like hosted a couple of dinners and stuff too, but it was very sort of very relationship-based. A lot of the relationships started through cold email and then would sort of maybe progress onto like a dinner or something. And it's because we were basically looking for design partners and we wanted companies to really close relationships with. And uh and on like on day one back in like 2022 there's also no like product really to begin with. And so you kind of need to have those relationships to gloss over the fact that like you know there's no product basically. [laughter]
Jake Aaron Villarreal: Yeah.
Tom Firth: And then we're now in a place where we've actually got a pretty good product. Like the product's like getting, really starting to become quite... I say this as a British person. We're always very measured about how we say things, but I think it's really starting to get pretty good, which in American probably means it's the best thing ever. And uh [laughter] and so we we are actually just starting the last few months to lean more into kind of the marketing side of it. You can sign up for an account like self-serve now, for example. And most of our customer base is still sort of these more enterprise style arrangements, but but but we are looking to open it up. And as we do that I think we are starting to shift to honestly stuff like this, more kind of like founder content, all that sort of stuff. And so over the next year or so I think expect to see a lot more of this kind of thing. But but yeah I don't know, cold email has been good as well. LinkedIn too, same kind of categories as email for me. Just messaging people cold on LinkedIn is surprisingly effective for like a lot of stuff in life. It's it's amazing what you can get by just asking someone on LinkedIn.
Jake Aaron Villarreal: Yeah, I love the strategy of the design partners because we've actually been approached as well to "Would you try this? Could you help us define what you want and what how you operate?" And yeah, you know, and it's really if you can find the right ones, it's it's a good win-win, but also a long-term potential customers and also introductions to their friends and their their network. So, I think it's a really good strategy. I I know that I don't know if it's Y Combinator or some of the other organizations that really promote, you know, start building with what your customers' needs are and have them share with you and have partnerships, which is yeah, like a friendly way to do it.
Tom Firth: I think so. Yeah, I mean, YC definitely promote it. But it it's almost intuitive I think in B2B space. Like it's like, like the minute you seriously go and try some, try and sell something to a... I built, I had to run with the example as as as we mentioned them already. They're not like a big company, but it's 5-600 employees now. Like they've raised money at a over 10 billion valuation. I can't I can't remember exactly what it was the last round they did, but it's like a it's a big company and it's growing fast and it's very mature in some ways. And the experience of like trying to sell something to an organization like that, let alone an even bigger organization, before you've got a product is yeah, is interesting. You kind of have to do it. If your goal is to sell to companies like that, then you kind of get forced down this design partnership route almost because because either you've got a very, very mature product that that ticks all of the boxes like the enterprise single sign-ons and the SOC 2 certificates, and we have got all of those things now, but you don't have them on day one. And basically consulting can like close the gap. And you make, you you start, you start these arrangements that look a lot more like consulting relationships with a, but with a sort of subscription fee attached and they kind of go pretty naturally from there. And I think it's the way to do it. It just takes so long to build like B2B products that you have to start somewhere on day one I think, and yeah, the design partnership works.
Jake Aaron Villarreal: Yeah. Well you've been able to do it and you've also kept your team lean and mean. What's worked for you in terms of assessing talent and making sure you've got the right people on the bus and as you look to go bigger at some point like what's what's that strategy look like for you?
Tom Firth: Yeah, we okay, we've been... hiring is something that like I think we've done a very good job of and I won't claim it's entirely down to skill to be honest because [laughter] because because like but both me and I, my co-founder, had done a fair bit of hiring before, but it's very different doing it for this kind of role. Like if you if you've if you've been a hiring manager or something in a company before, then it typically you've got some relatively narrow scope as to what it is. You've kind of like a lot of the parameters are filled in for you like how much money you've got to spend and exactly like what the role this person needs to be for and how many of them you're allowed. And and it's it's all just so much more open-ended when you're starting your own thing and trying to find someone for that.
And the thing that we indexed on pretty hard, and which I think was a really good call, is just get people that are in general smart. I think experience doesn't really matter that much for it. There's a handful of roles where that's like like you want a bit of experience, but by and large it's basically are they really smart? Are they really easy to work with? I think that's super important. I'm not saying they should be like a pushover or something, but they've got, you've got to like want to look forward to coming to work with that person because you're going to spend like the next two years in the office with them with like five people in a room potentially. And so you better like them. I think that matters.
And then we always kind of use this word like "hustle", which is a bit of a vague word, but there's this general sort of bias to action. I think that's very, very important. I think people that pontificate and you know, I'm not like anti-intellectual. In fact, it's the opposite. Like I like thinking a lot, but I think you do need a very, very strong bias to just go and get stuff done in in these companies. The problem is rarely like "we didn't think hard enough about it" and more that "we're just missing information." And so you kind of want to be out there in the world just like trying stuff. And some people find that very hard and some people find it really easy. And I think you do want to skew towards people like that. So yeah, I don't know like it's it's really hard. It takes a long time. The first person we hired, I think we we interviewed just... I can't remember how many. We looked at over a thousand resumes. It took it took a long time. And the first we hired out of network... that is in network as in it's somebody you've already worked with and you know, I would strongly recommend that if you can for sure. That's we've had a couple of really good people that way, but you do reach the end of the network sometimes and that's when it becomes like more interesting or a much harder game.
Jake Aaron Villarreal: Yeah. Yeah. We're usually brought in on those edge cases where it's like, "Hey, we've hit our bottlenecks. We've got the funding. Our investors are saying, 'Hey, are we going to build, grow, scale, like what's the what's the bottleneck?'" Because every company has, you know, a bottleneck or...
Tom Firth: Yeah.
Jake Aaron Villarreal: ...throat choke that's holding them back from really building and scaling. And sometimes it's just evolving as a leader, too. Like you have to also grow to the next level your company needs to go and not manage where you're at today, but you know, where are you going to be headed? Because you have to really operate two companies. The one you have and the one you're going to be. So, as as a leader in your experience so far, where have you evolved? Where have you seen growth in your in your own self?
Tom Firth: Yeah. I I hope I hope I hope if you ask my wife or my colleagues, they would they would say I've grown a lot because it certainly feels like I have and I've certainly worked hard at it. I think there's like three main areas potentially. One is definitely kind of self-awareness. Another is sort of resiliency and another is kind of basically like letting go of things. And I I'll go into a little bit of detail on those. But I think at a high level like I spend quite a lot of time thinking about this sort of stuff because the job does put you face to face with your own shortcomings like more or less constantly. And so I've got like all these Obsidian docs where I like analyze my own failings like all the time. It's very embarrassing. [laughter]
But but like the self-awareness thing I think is important because you basically you are going to make mistakes. I think is just the truth. Even if, you know, if I were to do this again, I'm sure that I think there's things I would be able to do better. But but but like even then I'm sure there'll be mistakes. There'll be mistakes I make. Like some of the things I thought I did wrong may turn out to be wrong and not the way I thought. And you just have to be really, really kind of like aware of the mistakes you're making. Recognize them quickly when you make them and and then and and then act on them afterwards as well. And I think if you haven't got this kind of like muscle for just looking at looking at yourself in the mirror and kind of like reflecting on it, then then I think that's a pretty risky spot to be in. And you can't always depend on other people to tell you. I've noticed like it's yeah, sometimes people aren't looking at the same way as you. And sometimes people are just nice. I think it it's it's like people don't want to tell people that like they're bad at something or that [sighs] you know. So so so I think I think you do have to depend on yourself for that.
The resiliency thing is like kind of related, but again, yeah, you're just gonna make mistakes. Things do go wrong. And I think you have to be the sort of like a sort of rock of optimism in the company. Like no one's going to believe in this thing anymore than you are. And if you kind of give up on it and like feel despair, then what can you expect of like the people you've hired? So I think you do need the ability to like no matter what happens, you've just got to like keep believing in it and keep projecting confidence in it. And yeah, that's very important.
And then, you know, and then the kind of what's the third thing I said? Uh, uh, oh [clears throat] yeah, letting go of things. Like, man, that's like the hardest one in some ways. I think most people that get into this are relatively competent or at least think they're competent. We'll say that because whether you are competent, whether you think you're competent, they both they both result in you thinking that the way you think you do things is like the good way to do things. And if you hire people and they're like capable, they also think the same way. And guess what? You're not always going to get, you're not always going to agree like you'll be like "I want it this way" and they'll be like "no I want it this way." And at that point you have to decide like "am I going to really stick my nose in and decide how every little thing happens and or am I going to just kind of hire smart people and let them get on with it?" And one is just like not scalable basically. You can... like I don't think I believe in founders just fully sort of stepping back, especially in like the early days at small scales. But I do think you need to find leverage on your time and and and yeah, like getting smart people to do things for you. It's basically that leverage. So, I think you got to let go. It's hard, but and I find that hard still, but but you do.
Jake Aaron Villarreal: Yeah. Yeah. It's one of my biggest challenges is is is leveraging others to do the things that you don't have the time to do because then you become that bottleneck or the chokehold of the company. And delegating and trusting and evolving and leading your people and also modeling how you want them to be. Like all those things you you had to think through and figure out. And I know in our first acquisition that we went through you know we built the company and I had two co-founders and when we ended up actually getting acquired there was a lot of self-reflection on what did we do right? What do we do wrong? What was the value of going through the process? And it was a great outcome, but you're never going to get it all right. That's kind of the beauty of the journey of the startup world [snorts] is you know you don't have that crystal ball and that road map to to make it all the time the right decisions. But, you know, I think people are aware of that and you just do do the best you can and hopefully you stack up more wins than losses and if you do then...
Tom Firth: Yeah.
Jake Aaron Villarreal: You keep going.
Tom Firth: Exactly. Yeah. Exactly. And I think that is kind of is the game. Like I think if you don't if you don't like that, if you don't like the uncertainty and you don't like this kind of process of understanding it, then I think you're not going to like starting a company because that because that is the journey. And some people for whatever reason just like that like get a kick out of it. Like I don't I I struggle to imagine myself going back to like a regular job having having done... I just don't think I you know there's there's something about building a company and leading teams but also building [clears throat] progress and momentum in a company and part of that is you know what are you doing day-to-day or week-to-week, month-to-month, quarter-to-quarter of moving things forward and showing progress that ends up becoming satisfying and then you build momentum.
Jake Aaron Villarreal: Yeah. What's your mindset when it comes to just the attitude of building a company because...
Tom Firth: Yeah.
Jake Aaron Villarreal: You want to keep it exciting and also positive.
Tom Firth: Yeah.
Jake Aaron Villarreal: And there's a lot of ups and downs in the process, but you're also leading a team.
Tom Firth: Yeah.
Jake Aaron Villarreal: And they're looking to you whether it's on a visual call like this where they can see your energy or it might just be the email that you send them and it impacts people.
Tom Firth: Yeah.
Jake Aaron Villarreal: Like what, what's your thoughts around leading and building a positive vibe and culture within the company?
Tom Firth: Yeah. Yeah. Yeah. This is a good question and it's very hard. Like I think this, this is arguably a tangent, but this is definitely something where as like a Brit building a team in the US that's been like particularly challenging because like cultural differences will sometimes make me like do stuff that in my mind means one thing and in the team's mind means another thing. And I've had to become like quite aware of that. And I've been practicing my hyperbole, but I'm not there yet. [snorts] You know, [laughter] you basically you do, you have to strike this balance between like, like one of the easiest things you could do arguably in the short run is basically just shout really, really loudly about how like great everything is. The kind of like overpromise like methodology. And I think that's tough. Like I I my experience has been you don't want to go like that way per se because things do go wrong and it's harder to deal with kind of like the the the kind of like down part of the ups and downs if everything has been painted as this kind of like rosy sort of like everything's perfect view. So I think you do need a very kind of like transparent and open take.
And I think you know what's worked for us so far is basically it's it's kind of like taking this how do I put this... like the reflecting on the state of the business and where we're going is something that we do pretty openly with the whole team and kind of invite everyone to join into. And I think everybody feels that they have this shared stake in this kind of thing that we're building and they know that the goal is to like, you know, we've got a sort of long run objective in mind, a thing we want to build. And the objective with the work is to sort of is to troubleshoot and figure out like what's the path from here to there. And we set these kind of like, you know, "Hey, we're going to run the next quarter like this initiative to try and figure out like this based on what we saw last quarter. Like we think it's going to work, so X, Y, and Z. But if it doesn't, it just means we kind of like analyzed wrong and so we'll take the data again and like reassess." And so I've made it quite like process-oriented I guess you could say and like focusing on the kind of like journey and the process rather than the outcomes per se.
And then when you do get good outcomes, you just have to like celebrate them and and yeah, lots of loud public praise when things do go right. And and then well I I forget exactly what the quote is but and who it's from, but there's this well-known quote somewhere about praising sort of when you give praise kind of giving it like to individuals, but when stuff goes wrong and there's like criticism quote unquote, it's like never aimed at individuals. It's always aimed at like the the process basically. And so I think you need to kind of balance that like open culture of discussing what's gone wrong and what's gone right and how do we course correct and making that just an ordinary part of like everyday dialogue combined with this kind of like like "so-and-so crushed it on this like great work," "Oh, we met the deadline. Let's go and like have dinner tonight" or something and all this kind of stuff. So I don't know kind of how we've... everyone seems positive so far. They accuse me of being the most negative person on the team. today. [laughter]
Jake Aaron Villarreal: Well, that's great. Yeah. [laughter] No, I I I I hear where you're coming from and I I like that. Yeah, we do the, you know, praise when things are done well and if there's issues, you know, sometimes it's more just individual like you take it offline and "here's what we're seeing, here's what we want to see done differently." So, they're feeling good about where they're at and that can also get better. What's, what's as if you look at the company today, what are you excited about? What's on the roadmap as you look into 2026?
Tom Firth: Yeah. Well, I mentioned already that the kind of like the largest change I think of the last couple of months is that we're sort of transitioning from this like pseudo stealth. We we never did like the stealth mode thing, but from this sort of pseudo stealth mode working with design partners to like just having an open anyone can sign up platform. And 2026 is just all about that. We've got the kind of core of a product that we're really proud of and that we think just absolutely rocks and there's lots of things to improve in it still, but it's ready to like open up to like all sorts of people. And so 2026 is basically all guns blazing on the kind of marketing and that kind of stuff to to get owner folks in to use it and keep refining it for these more sort of self-serve use cases.
The product roadmap is, I mentioned already there's really kind of just like two sides to what we do. Like one is the data infrastructure and the other is the kind of like the tools, like the actions I guess you can do within it. And so the product roadmap really like orients around these. It's just making it really really easy for people to plug in their data sets and really really easy to plug in all the tools, all the kind of actions that you can take within their within their business. And and yeah like it's it's interesting because for the first time ever in this business it just feels like the roadmap for the next two years is just very very clear like what it should be. Like I think we really know what we're working on and we're all jazzed about it and we're just pushing hard in that direction. Whereas two years ago or like 18 months ago, it was still very much kind of like in the wilderness like "what exactly is it going to be?" And I think we know what it's going to be now. It's going to be, it's going to be cool.
Jake Aaron Villarreal: Yeah. Well, we talked a little bit about, you know, as a as a hiring company, like we're on the recruitment side and we we see this a lot from a bird's eye view of what, you know, like the thousand startups in AI that are launching every month, what they're looking to do, where they're at, and you know, these kind of hybrid roles where it's defined as a forward deployed engineer is part of your company and then on site or even remotely connected with the customer. Especially, especially in, I I would think design partnering would probably be a really good option to have that type of person there to really capture and bridge the gap between what the customer wants and what they're sharing and what they're actually, what you're hearing on site and then bringing it back to the product team. What's been your success around hiring these people? Because forward deployed engineers have many titles: solution engineers, post-sales, you know, account manager, they're out there. They just have different titles and they have different core skill sets. But for you, what's been, what's been an area that's helped you nail that down?
Tom Firth: Yeah, this is a... So, I'm I'm, you almost certainly know more about this than me. So, it'll be interesting to hear what you think of my my take on it, but I I think the biggest advantage we had in hiring, so we've only got two FDEs currently, and then arguably a third or a third and fourth, which is me and my co-founder. My co-founder Ibe is sort of like the original FDE in the company I would say. And I would say the reason we're like, we had a nose for finding people to do it was because of him. Basically the job he did at Peloton was in many ways basically an FDE job but sort of like deployed within Peloton when it grew that it had that like crazy couple of years where it just grew.
Jake Aaron Villarreal: So it was insane. Yeah. Yeah.
Tom Firth: The stories he's got from that time are just absolutely nuts. The the rapid rise and equally rapid fall were kind of crazy [laughter] but...
Jake Aaron Villarreal: Yeah, I know that too. Yeah. Yeah.
Tom Firth: But I think the FDE role, it does depend on what your product is, but our experience is that it's it's a good chunk technical at least for our product, but it's much much more about really understanding what the like business outcome is for the the like person you're working with and trying to cross that bridge between or or bridge the gap between, you know, their world and their sort of terminology and their goals and the specifics of what your product and platform can do. And my experience has been that like it's that like translation layer that's sort of the bit that is hard to find. And we found...
Jake Aaron Villarreal: Yeah.
Tom Firth: Basically what we've done is the two people we've hired that do it, they actually both had fairly technical backgrounds. One was a bit more experienced. She worked as like a data scientist before. And another one was actually like a fresh grad. She she she did a she did a computer science like degree but had done in some interning for us over a couple of summers. And so we knew she kind of like kind of got the people side of it too. And they've both been incredible. And I think it's because they can sit down on a call like a normal human being and have like a very, you know, normal conversation with somebody who's not technical at all at the customer and just understand like what it is that they want and then they can come back and they can have a conversation with me or an engineer on the team and be like "This is what we need. I think we can like do it like X, but I'm not sure about this bit or like this bit of the product might be like, you know, not there yet." And then yeah, basically kind of like have that conversation with both sides to make it meet in the middle.
So yeah, in terms of like specific roles, I don't know if I have a good answer for like if I were to run a screen on LinkedIn for like people with job titles that would be good at FDE. I guess data scientist might be one interesting one but I think it's much more about the character, the kind of person. Like I think you do need to be a bit technical but I think you need to be mostly just like good at understanding business problems then mapping those back into technical stuff which yeah lots of engineers are really bad at that and and surprisingly some people who are not engineers are quite good at it so yeah.
Jake Aaron Villarreal: That's great, that's really, I mean a role that I know a lot of companies are hiring for today. You know, it's you got to find that person that's got the that dual mindset on the business front, on the technical front, product-oriented, and a great communicator. I think if you have all those, you're going to hit it hit it on the head.
Tom Firth: Yeah, that's exactly right. Like I wouldn't be surprised if you found some... like I wouldn't be surprised if like somebody who's done a couple of years at McKinsey having done a computer science degree or something like that. Like someone like that is probably really good at it. And I think a lot of data scientists...
Jake Aaron Villarreal: Yeah.
Tom Firth: ...also good at it. It's an interesting one. It's an interesting one. There's no, there's no degree called FDE that you can like [laughter]
Jake Aaron Villarreal: Not yet. Anyway, well, you you've got the right setup on your team as you head into next year. Super excited to see how things continue to go for you. For companies out there, businesses that are looking to build agents or automate parts of their business with AI, Tom, where do they go to find you? Where do they go to find your company?
Tom Firth: Yeah, pretty straightforward. The best place is uh our website, cotera.co. You can go there. You can, there's contact information on there. So, you can either reach out or you can just sign up and try it out. Just hit the sign up button in the top right of the page. Or yeah, feel free to reach out on LinkedIn. It's Tom Firth. Just search Tom Firth Cotera I guess on LinkedIn and it should show up. The picture on LinkedIn is the same as the biopic I sent you for this podcast. That should be easy to identify. Yeah, feel free to reach out any time. [snorts]
Jake Aaron Villarreal: Okay, perfect. Well, thanks Tom for joining on today and I appreciate your time and also appreciate the time of the listeners for spending it with us to, with us today. It means a lot to me. I'm your host Jake Aaron Villarreal signing off for now. We can't wait to catch up with you all on the next episode. Until then, Tom, the world, take care. If you like what we're doing, don't forget to subscribe, leave a review on Apple Podcasts or wherever you listen, and follow us on YouTube where we go behind the scenes to learn what it takes to be a startup founder.
.jpg)
.webp)
.webp)
.webp)
.webp)
