What Breaks First, Episode 5: Chris Wyman (Chief of Staff, Deeptune (Mercor)) with Ian Myers (Founder & CEO, Oceans Talent)

[00:00] Chris Wyman (cold open):
there can probably only be like one big thing that we are tackling at any given time because if we don't figure that out, the business is dead.

[00:10] Ian Myers:
All right, I am here with Chris Wyman, Chief of Staff at Deeptune. Chris, thank you so much for coming on. I'm wondering if you could share a little bit about yourself, your background, and what you do.

[00:22] Chris Wyman:
Totally, I'm Chris Wyman. I'm a startup operator. I've spent the past 10 years at a bunch of companies, growing very quickly, some of them smaller, some of them later stage. But currently Chief of Staff at Deeptune. I had to start at building reinforcement learning environments, essentially to teach the AI models that will get released in six months from now, how to do the work that you do in your day-to-day life. We were recently acquired by Mercor this summer, have been on the team since last June, was the first non-technical hire on the team.

[00:51] Ian Myers:
Wow, first non-technical hire last June, so that's a year.

[00:56] Chris Wyman:
Yeah, just over a year.

[00:57] Ian Myers:
To acquisition, short span of time. I also noticed that you mentioned rapid growth and quick scaling of businesses. You were at Gorillas and Uber Eats?

[01:10] Chris Wyman:
Yep, I was at Uber for about a year and a half on either side of the IPO, and saw a lot of craziness in that time period. And then left to join Gorillas to help launch the US market. This was at a time where essentially interest rates were at zero.

[01:27] Ian Myers:
I remember that.

[01:27] Chris Wyman:
And venture dollars were unlimited, especially in the delivery space coming out of COVID. So that was a crazy, crazy operation to launch the US, grow the team very quickly, reduce the size of the team, grow the team again, go through a lot of ups and downs. And so I have a bit of a internal magnet to the most chaotic situations that I can find myself in. Otherwise I'll get bored.

[01:51] Ian Myers:
Excellent, excellent thing for a chief of staff to have. Just curious, I think food delivery, ride sharing, and AI are some of the places that have grown the quickest, gotten the most investment, blown up the fastest over the past decade or so. How does AI compare to the delivery food service space? Is it like totally new or are you seeing repeatable patterns in how things grow?

[02:20] Chris Wyman:
Well, I think it's really interesting, especially in this AI training space, because fundamentally it's a very human business. It's a marketplace of human experts across every corner of the economy who are lending their expertise to train AI models. And similarly in the delivery marketplace, it's a marketplace of folks who are coming together to get things from point A to point B. And so at the end of the day, you have a lot of very similar people problems.

[02:48] Ian Myers:
People problems. And I've heard that's your area of expertise. You recently wrote a post about the demanding requirements of hiring how many engineers was it?

[03:00] Chris Wyman:
You know, I would love to hire another 35 engineers over the next six months. We're a little bit behind target. It's really hard to find great folks. It's a weird market that we're in today, but with the nature of work changing, I think that more than ever, having really strong judgment and taste is something that you just can't get in a lot of folks. And it's not something that you can completely outsource to an LLM.

[03:24] Ian Myers:
Hiring is of course one of the most crucial skills any operator can have. I am curious as a lifelong operator yourself, did you engineer a like move to AI? Like, were you looking for that to jump to when you went to Deeptune? Are you just kind of in it with the latest, greatest industries? Or was it something that you kind of fell into by accident?

[03:48] Chris Wyman:
I think it's a mix of both. So before Deeptune, I was at an early stage dating app for about 18 months running operations and marketing. We wound down operations and I knew that I didn't want to jump into the next thing right away. So I took a sabbatical, which was incredible. I spent the first six months kind of doing nothing but enjoying life. And then the next six months doing a bit of soul searching and figuring out where I wanted to land next. I had this hypothesis that a chief of staff role might be a good fit because I've always enjoyed two things. One is a lot of breadth in the role and essentially not doing the same thing over and over again, but instead kind of having my finger in every pie in the business. And then second is having a really close relationship with an executive or a leadership team and being kind of a trusted confidant. And so the chief of staff role really blends those. I went and basically chatted with about 50 chiefs of staffs at a bunch of different types of organizations, mostly early stage, to pressure test this hypothesis of what is it actually like and what do they like about it? What do they hate about it? What red flags should I look out for? I distinctly remember when I was looking for the, like where do I want to be a chief of staff? I knew that I wanted to be early stage and I was at the gym and I saw this article or maybe I was listening to a podcast and it was a stat from, I guess this must've been Q1 of 2025 where something like a quarter or more of venture dollars were going into AI startups. I was like, all right, well, that narrows the search a little bit because you gotta follow the money. Those are the companies that are hiring and ultimately what is gonna land. But I do nothing about the human data or AI training space, but it's a really interesting space that most people don't know a ton about.

[05:36] Ian Myers:
Despite a lot of coverage, I think it is. It's very surface level knowledge, but I'm assuming you had to come up to speed pretty quickly. How was that transition into an environment you knew? I think you can intuitively grasp food delivery and you can kind of like intuitively grasp what ride sharing is. It is probably mentally more challenging to grapple with what Mercor is doing, what Deeptune is doing. So how you seem to have been very strategic about getting yourself into these positions. Usually when people decide they wanna take on a new career, they read an article and then apply to jobs. They don't spend time actually talking to 50 people about whether or not they should do it. So what was it like getting started in two very unfamiliar ways?

[06:25] Chris Wyman:
Yeah, I mean, I essentially got thrown right into it. So when I joined the team, it was Tim, the founder, one engineer that had been there two weeks and another that started on the same day. And we had this pilot with a frontier lab that was great, but we didn't know how to deliver it. We kind of knew what the data needed to look like and what needed to be done, but we hadn't finished anything and we needed to essentially do like a handful of these environments and create the data sets and really spent the first few months just figuring out all of the things that we didn't know. There were parts of the process that felt very similar to things that I had done in the past. So data annotation was very, and hiring data annotators, managing this human workforce and a very operationally complex process is very similar to things that I've done scaling customer service teams of you find folks that can do this work, you create a playbook, you go find markets where there's some kind of arbitrage in the cost of talent and you like scale that workforce and you like figure out what metrics to monitor and how to scale that. There was like things that I could draw on there, but the actual work that we were doing of build a simulation environment to replicate like the work of an accountant. Like I knew nothing about accounting nor reinforcement learning. So a lot had to be learned on the go.

[07:51] Ian Myers:
I mean, you have talked a lot about hiring and setting up teams and standing up processes. What were you expecting to be the most difficult thing when you started the job and what actually turned out to be the most difficult thing? Were they the same or was it something else entirely?

[08:08] Chris Wyman:
I think that the data space is interesting because one, it's very, very new and two, the buyers are very concentrated among a handful of very large frontier labs and large tech companies. And when I had talked to other seed stage startups, essentially all of these founders were figuring out, okay, we have a semblance of a product. We have maybe a design partner and a handful of customers and we need to figure out both how to scale distribution and how to actually make this product work and how to build a team around it. And how do we go from like zero to maybe a million or 2 million in annual revenue. The companies in the human data space are doing like orders of magnitude more before raising a series A. And there's an insane amount of demand and we are in a very supply constrained market in that there's truly no way to produce more data or enough data for the labs to consume at any given time. And so this problem of how do you sell it or how do you scale distribution is not actually what needs to get done. You need to figure out how do we produce data at the scale that the OpenAIs and Anthropics of the world demand. And so it's much more of like a technical problem, a scaling problem, a people problem of how do we get enough scale so that we can actually be a serious player in this market.

[09:37] Ian Myers:
So how do you turn everything into a volume game really quickly? And the scale, to put it into perspective, I think it's important to think that like, oh yeah, start a scale quickly. They have the funding to do it. They can spend more than they make. They can go from one employee to 100 employees to 500 employees in like three to four years. It's a hugely compressed timeline. For a lot of these companies, it's that timeline in like six months, right? And so the volume, like everything becomes volume. Onboarding becomes volume, payroll becomes volume, you know, Slack becomes volume, email becomes volume. How do you think people should think about that kind of rapid growth from the perspective of prioritization of business needs as you're growing that quickly? Like what was the most important thing and the least important thing as you were going through that scale?

[10:31] Chris Wyman:
Yeah, I mean, I think there are really two things that I think about. And as an operator and somebody with a bit of a perfectionist tendency sometimes, like I love finding a solution that has a really long shelf life. Like we have figured out this process and it is going to work for a long time. At the pace that we are scaling, one, there can probably only be like one big thing that we are tackling at any given time because if we don't figure that out, the business is dead. And if we try to do everything else, we're just wasting time. The second is like being really comfortable figuring out a solution that has a shelf life just long enough to give you leverage so that you can move on to the next most important problem. And sometimes that solution is only gonna work for like six weeks. Like when we start interviewing someone, there is like an acute need in the business and in four to six weeks when they start, like we've probably solved that and there's like 10 other issues to solve for. And being an operator in a company like this is exactly the same thing where like the most important thing today in a month will feel trivial because there are five other things that have popped up. And so it's a lot of figuring out how do I solve this in the simplest, fastest way so that I can move on to like the next most important problem.

[11:50] Ian Myers:
Got it. The idea that working on a solution for something that's not even gonna be a problem by the time you find the solution is such a horrifying thought. And I am enjoying my nice bootstrap services business, but you know, there's pluses and negatives to both. I wanna shift gears quickly to focus on kind of the two quick, big things that must have happened in your life over the last year. So basically you joined four people, five people, something like that. Hopes and dreams, already product markets fit, scaling rapidly, and then you raised your series A 43 million dollars, that's a ton of money. What were you doing, thinking, and feeling as the news of the raise was going out? Was it like, great, we have more money to hire more people to do the stuff I need? What did you like personally, what did you feel like when that happened?

[12:46] Chris Wyman:
It's really interesting because of the timeline of fundraising. So by the time that the press release goes live, the money is in the bank account and is actively being spent. And so that idea of like, everyone else is obviously hearing about it, it's a huge public validation of like, you have something to share with your friends and family, and people kinda know a little bit about what Deeptune is. But it, at the end of the day, sometimes actually feels like a distraction of like, this actually is not the work. The work is not fundraising, the work is not announcing the fundraise or responding on LinkedIn to everyone saying like, congrats, because the money's been in the bank account for a few months, and now we're expected to spend it and grow the business. And then we had our team offsite the week afterwards to take everyone to Austin and do some bonding because we had tripled headcount in a matter of months. And so it's like this really weird dichotomy of like, oh, we're kind of like this more grown up company, but also so many things are messy and moving quickly and trying to balance all of it at the same time.

[13:48] Ian Myers:
So you felt as though it was like, great, the news is great, it's validating, it's affirming from a third party perspective, but you're already back to doing the work. Like you saw the signature go out months ago and to no fanfare whatsoever, and then you kinda get back to work. Yeah, I think that's something people don't often think about, especially if they're, they haven't been part of a fundraise before. And then a couple of months later, you get acquired. Did you think, oh man, we might get acquired when you did the raise, when the raise was going on?

[14:27] Chris Wyman:
But the industry that we're in, you can compare a month to maybe a year. And so the timelines, while compressed, actually feel so much longer. And so much happens kind of, we live week by week in terms of every model release and every announcement and every change that's happening in the world. And so a few months really feels like a lifetime. Like I was telling someone, oh yeah, like last April when we were in Austin for the offsite, and I was like, and that was five months ago. But it feels like the longest five months of my life.

[15:05] Ian Myers:
Remember two years ago when we had that offsite and yeah, that's, I can imagine it feels like that, especially if you're working long hours, long days. And when you learned that the company was getting acquired, whenever that was, were you feeling like, oh man, this is incredible news. Or because you're the guy running operations, you were like, oh my God, my life just got way more complicated and annoying. Were you like excited, nervous, relieved, nervous?

[15:33] Chris Wyman:
I think I am somebody, regardless if it is work or making dinner plans, that immediately starts thinking about like the logistics of that and the work that needs to be done. And oftentimes like as a chief of staff, it is sometimes like kind of like being a duck and you think about the waterline and like the duck above the waterline looks very serene. It's like very rapidly kicking its feet. Out of the chief of staff often is like that duck, essentially taking care of like all the stuff that's below the waterline that probably the rest of the team isn't thinking about or the principal isn't thinking about because they have other things to be working on. And so immediately started thinking about essentially like everything that needed to get done leading up to the transaction closing. And then afterwards, it was a very interesting position in that there was a very small circle, essentially like two people that know about this deal and then a bunch of like well-paid lawyers. And you can't really talk about it and you can't really make decisions based off of it until it's done. And so essentially it's holding everything in your mind of, okay, what do we do as if this is not happening? And what is everything that we need to do knowing that it is happening? And most of the time that works out, but it is essentially a massive exercise in contingency planning and thinking through kind of every scenario and every checklist of things that will need to be done and having that ready for as soon as like pen hits the paper and we can start executing on some of the integration stuff.

[17:02] Ian Myers:
How do you organize all that just out of curiosity? Like what's your stack of stuff that keeps you on track of all of this craziness?

[17:10] Chris Wyman:
I'm a Claude Code power user. I'm using Claude Code out of Obsidian. I use it as like my personal operating system, essentially all of the context of my job. And it's connected to like all the Slack and email and other data sources in the company. But it's that and a lot of like pressing record on Granola and talking to myself for 20 minutes before I go to bed to try to get all of the thoughts that are keeping me awake out so that I can enjoy some peaceful sleep before I do it all over again. But essentially it's like, how can I get as much context down as quickly as possible and then apply the judgment of like, what is the thing to do next? That's like all you can really control. It's like, what is the next most important thing and what do I need to do to unblock it? And like everything else will eventually happen or maybe it'll sit on your to-do list and it's no longer relevant by the time you get to it.

[18:01] Ian Myers:
Yeah, so you're not surprisingly a Claude power user. I am curious, I mean, from where I'm sitting, you work at an AI company, you use AI daily and constantly, you're surrounded by the industry that is AI and you also hired a person from us to help with that. So why, I guess is my question.

[18:32] Chris Wyman:
We just are not at a place where AI can do everything for you. But even if it can, you ultimately need somebody that is accountable for the outcomes or accountable for the work. Like if there's a mistake, as much as I can type angry things to Claude or be frustrated with Claude or give feedback to Claude, that's not really accountability and that's not really having someone who owns it. And so there's no peace of mind in that. So it really is like, one, the systems or the tools today do not do everything for you. And two, nothing really beats having a person that you can trust to get the stuff that matters done to help grow the business.

[19:15] Ian Myers:
I have asked a lot of people this question and I have not heard from a lot of people their response to accountability. I also feel when we have people that hire executive assistants from us, I liken it to having a personal trainer. Like gyms exist and you can go to them, but you might not, but you might not work as hard. You might not see the results you want without someone there constantly keeping you accountable for what you're doing. And so I think the accountability piece of it really resonates with me personally. I'm curious how you divide the labor between, or how would you divide the labor between a human support system and an AI support system? Like what are the areas where one is excellent you rely on and the other is excellent you rely on?

[20:03] Chris Wyman:
Like tooling and like an AI system works really well when there is a more deterministic outcome that you could expect. Like create and process this invoice, complete this financial analysis, send this email, things that are templatized and repeatable and you know how to evaluate whether or not it was done well or not. And sometimes you need a person for this stuff or like you actually don't know what you want or you don't know what the solution is. You just know that there is a problem and that you don't have time to do it and that you need someone else to go figure that out for you. And like I think about, like we planned it offsite and it was so helpful to have, you know, I know that we need to do dinners and I know that we need to do like have lodging and have transportation and have activities. And I surely can go to ChatGPT or Claude and ask like, where should I eat? What should I do in Austin? But it's not at a point where it is going to like compile those options and compare against my goals for the offsite and the trade-offs that we're making of, do we do this or do that? And then actually like call up a business and make a reservation or move things around last minute because overall hungover and now lunch needs to come earlier because like people are dying at the house and like sometimes you just need a person to like jump in and help fix things.

[21:28] Ian Myers:
I agree with that and I also think that rapidly changing conditions are kind of difficult. There are plenty of instances where I've tried to get something done with AI and there's just like one small component that it doesn't quite understand or like doesn't intuit that a person would obviously think about. So it's sort of like the last mile of it all that really unravels things, I guess, unless you've got a hand on the wheel. In terms of planning the offsite, I'm curious, like what were the other things? I'm trying to paint a triangle of what you did day-to-day as an operator, what your support system Moneli did as a assistant to the operator and then what you relied on AI for. So were there like categories where you felt like I have to do this, where I can hand this off to AI, where I can hand this off to Moneli? Like how did you structure and organize that in your mind?

[22:25] Chris Wyman:
Yeah, I think it's really helpful, I think beyond just the offsite and like taking a step back of like being a chief of staff and having this like laundry list of stuff that you are responsible for and having an EA in the mix. It's like, it can be very helpful for the team to distinguish of like, come to Chris for X, Y, and Z more strategic things and like ABC more tactical administrative things go to Moneli. That one saves me a ton of time and two makes it easier to work with me because I'm not getting these like constant pings and context switching and also like, I think a better standing in the organization because stuff is getting done. When it comes to the offsite and where the split was, I really focused on one, what are like the overall thing that I want to accomplish through this offsite, which is bring people together, give them a forum to connect and build relationships. And so I was like a skeleton of like, okay, we need to go to Austin. We need to stay somewhere. We need to leave feeling like we have values and know each other. We fill that in with like Claude can come in and help parse that into a minute by minute schedule. Cool, there's, we're gonna do X, Y, and Z activities. It's gonna take this long. Here's how much buffer we need. Like, oh, we need more buffer. We need to move this around. Like it's gonna take this much travel time to get between things. Moneli can fill in with, okay, well, like we are going to dinner. Where are we gonna go to dinner? Like go help me find some good places that we can go out. They're gonna work for our size group on this day, this time, these requirements, whatever. And like our welcome dinner, for example, and every dinner that we did, I had Claude generate the seating chart. And it was like 25 people. So it's like, I want people to mix around and chat with people that they're not working with on a day-to-day basis. That is something that I don't need to do. Like, I'm not gonna go lay place cards and decide like this person should sit next to them. And Moneli doesn't have the context necessarily to like pair people together. And LLM is great for this task of, okay, here are the 25 people across these like five activities. Make sure that they never get paired with the same person. And here's what they're working on today. Like pair them with people that they're not working with. But like an LLM is great at that kind of logic-based activity.

[24:40] Ian Myers:
I wanna end on talking about just chief of staff in general, which is a role that I think has exploded in recent years in terms of things people want to do. I remember maybe four years ago or four and a half years ago, a friend of mine who had went to a great school, worked at a phenomenal private equity fund, called me up and was like, hey, I really wanna be a chief of staff at like a tech company. Can you help me? And I was like, what? But you work at this like great private equity fund. Like why do you wanna do this? And she was like, well, you know, it's like proximity to CEOs and I can really learn everything. And I think I also like, I can learn what makes a business grow. And like, it's the greatest opportunity for me to get hands-on experience across a huge variety of departments. And I had said, what are you talking about? 'Cause it was not a popular role at that time. And it was like very, you know, like the president having a chief of staff, sure. But, you know, what do you feel so far about the role of chief of staff and you know, how necessary or unnecessary it is in an organization?

[25:55] Chris Wyman:
I'm biased because I'm sitting in the seat. But I think that-

[25:59] Ian Myers:
You seem like a guy.

[26:00] Chris Wyman:
The nature of work is changing. There's a great chart, I think, from maybe McKinsey where it shows like the rows are essentially like what domain you work in. So marketing or sales or legal, finance, whatever. And then the columns are essentially the percentage of your time spent working across domains. So how much time does a marketer spend doing marketing versus finance or sales or something else? And the lines are becoming more and more blurred. Like people are not just sticking in their swim lane. You're expected to be this like 360 operator across the whole business and be able to do a little bit of everything. And two, in the age of AI, trust and judgment and taste are so valuable. And you can't delegate that to an LLM today. And so more and more, there is this need for somebody who can be dangerous across all these different domains. Probably not an expert in any of them. They spike in one or more. And two, like the ability to synthesize all of this and just make good decisions and hold information that sensitive or confidential or just like balance everything. And so I think that like more and more, we are seeing lots of chiefs of staffs in the workplace. We're also seeing very small founding teams. We're seeing people who are solo entrepreneurs who are doing a little bit of everything. And I think that it essentially is all part of that same trend of like the future of work is becoming much more multidisciplinary and you're expected to be much more of a generalist.

[27:29] Ian Myers:
What was your major in college?

[27:31] Chris Wyman:
I dropped out of college. I was looking for a day job that I could do while taking night classes in community college. And so thought that maybe a desk job, admin type job could be good and stumbled into an admin job at an early stage startup and was learning more there than I was in my first semester general ed classes. So dropped out of--

[27:51] Ian Myers:
So you went from admin, like pure admin to operations to chief of staff. Was that kind of the career trajectory for you?

[28:00] Chris Wyman:
Yeah, more or less a stop along the way in the customer experience world. But my first startup job essentially was scheduling and picking up lunch and all this stuff that needs to get done so that there can be like a happy office.

[28:14] Ian Myers:
I guess if you were talking to other chiefs of staff who are trying to navigate the same thing you're navigating being essentially the one neck to choke for many leaders struggling in a very difficult environment, what piece of advice would you give them on how to do their job successfully?

[28:32] Chris Wyman:
I think it's actually not about what you do at work. Not enough people pay attention to what they are doing with the other hours in the day to take care of themselves so that they can show up. For me, that means like, I need to sleep. I need to eat. I need to have a good gym routine. I need to have like a handful of people that I trust that I can vent or talk about or be a sounding board and feel like I have this fulfilled complete life that sustains me so that each day I can walk into like a fresh set of challenges and that will make or break you. Like I have like tons of ideas for how you can be more effective in your job as a chief of staff or as an operator. But if you don't have that like system outside of work to be an effective, healthy, stable person, like there's no tool that's gonna make up for that.

[29:24] Ian Myers:
It seems like a role that could easily suck you into 24/7 work mode unless you effectively push back. So I think that is excellent advice. Chris, thank you so much for coming on. We'll wrap it up here, but it seems like it's gonna be an exciting few years ahead for you.

[29:44] Chris Wyman:
Thank you so much. It was great chatting.