Melisa Tokmak (GM at Scale AI, ex Chief of Staff at Scale AI)

Hey everyone, this is Erik Torenberg and you're listening to Execs by On Deck. Execs is a show for founders, operators, and leaders who want to understand the playbooks, frameworks, and tactics that leading tech companies today have used to scale. Our guests are tech execs in key roles at top tech companies who share their hard-won, earned secrets on how to scale faster. Melissa Tokmak is a general manager at Scale AI after serving as chief of staff to Alexander Wang, the CEO. In this episode, we learn Melissa's tips for scaling culture. notably her recommendation to find and hire people who give a shit. We also ask her what's the cheapest staff really and chat about how Scale is building the data infrastructure layer for AI. So Melissa, for the audience who may not be as familiar, can you define what Scale is and how you came to join it?

Yeah, Scale is the data infrastructure layer for AI. You know, when I say that, it is a mouthful of words and not exactly you understand at a first glance what it does. But the way to think about it is in order to actually operationalize machine learning and actually use it in the products and in the problems that we are facing day to day, there's a lot that needs to happen from building those models, but also even more importantly, feeding the right high-quality data to those models, monitoring and continuously improving those models across the board. So there's a lot of steps. You can think of Scale as that data infrastructure layer to be able to make machine learning operational in the real world rather than just an idea in the research projects or something that we're excited about. And I came to join Scale in a very interesting way. After school, I studied computer science at Stanford. I joined Facebook to be a product manager in new product monetization. I really did love my experience there, but I was craving for something more where I could really give it all I got and really work hard to build things that I actually believe are needed and longed for in the world. And I was about to actually take the plunge to build my own company. And around that time, very serendipitously, Alex, our CEO of Scale, and I met through an introduction. And it was not even a recruiting introduction. I was not thinking about joining a company. It was more about talking and riffing about machine learning and how it can be applied in some of these use cases in big tech. And at the time, in the teams that I was leading at Facebook, this was a big need. What Scale was doing was a big need. So he was talking to me about those ideas. We met and that meeting was, I mean, one word for it would be explosive. Really, I think we were so aligned in terms of our values, but also the way we look at what work is and what work should mean and how can people approach it. The passion about machine learning, but not only as a field, but actually its potential in the world if it can be actualized, right? So it was a great fit and we kept talking a few more times. And then about maybe a week or two later, he gave me a call and asked, that he had a role and he wanted to pitch me the role. And I was very excited to hear the pitch. And I can say, for many of you who maybe have met him or heard him speak, he's very good at understanding what people care about, I think, in their essence, and really hitting those points. He really hit those points for me in the pitch. So I was ready to kind of stop the things I was thinking about for myself and join SCaLE. Definitely, he was one of the biggest factors for that, but not the only one. I actually met more people in that week from SCaLE, not only in leadership that I was joining to, but everybody really in the company, right? And I had some friends from school working there. Everybody I met was extremely smart, but, you know, smartness, I think in Silicon Valley, like I did surround myself with incredibly smart people, Mm-hmm. but extremely dedicated and practical as well. Like, I was like, wow, these are the right people to build this business, which is a hard business that will require extreme dedication and hard work. And they'll be, you know, good business builders. They're practical people. They really care about what's the problem and how to solve it in the best way possible. So the rest is history. I think it took, maybe I joined in 10 days, something like that. It was crazy. Yeah.

Let's talk about the chief of staff role, because we've talked about how that means different things in different places. And you've been very deliberate in how you shaped the chief of staff role for what you wanted it to be and how you can make the biggest impact on the org. Talk a little bit about the role and how you shaped it.

Yeah, I didn't know much about chief of staff role. I mean, I could always see, right? I also had friends around me who have been chiefs of staff. But they have done it for mostly venture capital, right? It's very different for companies. So I did companies. I did not know much about it. The first thing, the role he pitched me was the role of chief of staff. But I think even at the time, he himself didn't know much about it. He just— I guess the thing that encouraged them is like, we're growing incredibly fast and I need to be able to scale myself and I need to be able to have somebody that I can really trust. to put on these things, and I don't have to think about it as if I'm doing it myself. That was really the guiding principle, I believe, for him. And for me, I talked with a few close friends about the role and like, what are the pitfalls? What are the things that make the role great? And consistently, one thing that came up is whatever you shape the role to be, really having that alignment early on and first. Right? And I think I was coming from a place of a little bit privilege in this case that I was not really looking for a job. I was not really looking to land it. So that gives you a little bit of a leverage in the sense that you can be honest about your needs, about your wants, and what you're hoping for that role to be. For me, I actually, I remember this, I sat down and wrote a document, you know, like, what do I want out of this? And for me, the most important goal in my life is To be able to start my own business, so I was like if I'm gonna do this, right? I'm doing this for this opportunity and for the people that I want to work with. I genuinely want to be around them to push myself to be better, right? Because I believe those people are better. So I structured it around that. Like what do I need to be able to build my business next, right? And that gives you a little bit. like very detailed points. For example, like, what is your role when you're talking to external customers, right? In a world of company, there's an external world and an internal world. What is your role internally? What kind of hats you can wear? What kind of teams you can build? So, every little detail about the role I wrote was around me being in the best position to see opportunities and start new business units in the company, including from the engineering perspective, product perspective, even sales perspective. Because I think when you are a CEO, you cannot just delegate that work. You have to be very good at being the salesperson and the builder together. So I wrote down these details and, you know, quite frankly, I was very granular, like very, very granular. Even the things I absolutely did not want to do, like roles or like, you know, the level of involvement I wanted, whether it's a board meeting or an external meeting, and the level of involvement I didn't want internally, maybe in the areas I didn't wanna do. And I remember before kind of I accepted in those conversations, he and I got dinner and it was a long dinner because I literally pulled this list up and I was like, we're gonna go over it one by one. And I wanna know, you know, if this expectation is aligning And if it's not, that's okay, but I wanna know it now and what your expectation is about that item so that I'm really coming into this with open eyes. And I'll be very happy. If I'm happy and working really hard, you'll be very happy. So I think that process worked for us, but that also worked because I always say the role of chief of staff depends on 2 other things. One, Where is the company at? Right? So is it at an earlier stage where like you genuinely need kind of a person that will extend the CEO in ways that this person will take on duties as if the CEO would? Like, is it you? Would you send this person to your most important meeting without you? Right? Would you let this person build a team? Would you let this person interview people? Like, do the most important interview? after they pass the technical skills for the culture interview. So like these kind of things, you have to understand where you are as a company and who you need, right? And can the person satisfy that? And also second is about the person you're becoming the chief of staff for. I'm sure this role is very different, right, when you're doing it for someone who has built companies 5 times, right? Or a first-time founder, or a younger founder versus someone a lot more experienced. I specifically was very interested in this because I respect people in my life if they are very smart and they work hard to get better every day. And I think Alex is the extreme example of it. And I was like, yes, I will love this. This person is very aligned with my values. But the second thing, it was because many things he was also just figuring it out. I think this is not only for him. Any leader in the world is just figuring out. But sometimes I think with a lot of multiple-time founders or very, very experienced founders, it might be that they already feel like they know the answer. I was less interested in that. I was interested in an environment where can we figure it out? Like, can I have that leeway and space? Because he's also figuring out, so like he'll be just open to great ideas, right? It's not immediately kind of maybe thinking like he has the right idea. He was very open to that feedback and just the meritocracy of ideas. That was important.

Totally. Let's talk about some of the main projects you did under Chief of Staff, and then we'll talk about your role after Chief of Staff. One of them was you helped shape culture at SCaLE. Talk about what that meant for you and what that meant for SCaLE and how you guys did that.

That's a very good question because shaping culture is very hard, right? We talk about this, but what does it even mean? Culture is shaped in the moments you don't see and you don't talk about, like when you're not there, right? So for me, when I came into SCaLE, it was early but not that early. So, first 100. So, they already have done an amazing job with those first people, really setting a culture that's absolutely amazing, right? In terms of the caliber of the people, the dedication to the work, the passion about machine learning, dedication to the customer, and just willingness to what we call run through walls, you know? So, I had something to be inspired by, right? It's not only like what I wanted it or like Alex wanted it. Shaped by those people who you bring early on. It's more about how can you make sure as you grow, you keep the important aspects of that culture, the things that make you unique as a company, all the while bringing in flexibility to bring others, right? So on that, I think maybe it's multifaceted. One, Everything starts, whatever culture you wanna set, you just have to be very intentional. And that means talking about it, right? Talking about and writing it down what you're trying to do. So we have really awesome, what we call credos, which are basically just values at scale that is about, you know, ambition shapes reality. For example, one of my favorite ones. The other one is run through walls, right? The other one is earn customer trust. You can be very deliberate about what are the things you want the company to demonstrate, but that doesn't end there, right? Usually you read these things if you don't refer to it a lot. So I really cared about building a mechanism where people in leadership are constantly demonstrating this, but also recognizing the people who are demonstrating this. Right? It is not unusual at all to get a message or an incredibly thoughtful and eloquent message and essay from Alex at like 3:00 AM because he worked all day on it, all weekend on it, and really to distill his message. And some of it he actually shares publicly as well, distill his message about the vision and really worked and finished and send it out. Or it wasn't uncommon at all that, you know, we would have customers and they would suddenly have a different need. But we do want to go above and beyond really earning that trust and delivering. So, you know, we would work through constantly as a team, like a big group of team would continue to work through to deliver that. It doesn't mean like everybody has to be like that, but really demonstrating that leaders can be like that. And with the effort, that effort, if it is demonstrated and that dedication is recognized, right, and rewarded. So that's definitely the one part of it I really care about. And the second thing is making sure that you're bringing and expanding the company with people who are carrying similar values, like in the things that you don't want to compromise on. What are those values and how can we make sure? So we really pay attention to our interview processes. And one thing that, you know, never changes across all roles is really that credo interview, what we call the values interview. We try to learn about what motivates people in life. It doesn't even have to, like, it's not about machine learning or it's not about other things, but who are you, like, as you and, like, what motivates you and how can we demonstrate from both sides that we're value-aligned so you'll be happy and you'll be thriving here and we will keep the best parts of our values and culture together. So.

Yeah, the post Hire People Who Give a Shit by Alex is a really fantastic encapsulation.

Yeah.

What you guys are trying to do. How do you manage people out who don't embody the culture?

Yeah, I think in my, right, like my teams, like I can just kind of be specific about that more like from my side. A few things. I think first of all, you need to understand, right, like why they're not a values alignment. So, what is going on there? It's not like you can just put a rubber stamp of not values-aligned, right? Like, it also means it's not about the person. It means, to be honest, you messed up. Like, you hired, right? Like, you put in so much effort and resources into hiring someone. So, you can't like— it's not a direct rubber stamp, but you have to look through it. Is it about like, does this person have the space and the things they need to thrive in here? Right? Like, that's important. Was it— is there really in the most important things we care about with ambition shapes reality, like hiring people who give a shit and, you know, ownership is the job, like stepping up and really taking on the ownership when things are good, but also when things are bad, right? Is there no alignment there? What are the examples of that no alignment? You have to first— and we genuinely really have a extremely transparent conversation and culture at scale where it is direct and transparent conversation, like one-on-one as well. So first, like, you would really deliver the feedback and talk about and work with the person if it's not going to work out. I haven't seen cases where after that, if nothing improves, it's that hard to manage people out because it's so clear it's both ways, right? If you're Not feeling if I'm not feeling the values fit here, it's also miserable for me, right? Like every day, like if I'm not really waking up, feeling encouraged and empowered by the way and the values that we're implementing here, it's it's also miserable for me. So I never seen a case where it's difficult after that. If nothing improves, it just becomes very mutual, right? The thing the most important thing. That I think any company can improve on, and definitely us, is, is finding the right and fastest moments that you can force that conversation. Because if you are not timely and you are continuously delayed in those conversations, you're not doing a service to either of the parties, right? First of all, for that person, like they don't have the room to thrive and they may not know what's going on or they know, but like nobody's bringing it up. And second, for us, we are a startup. That means— and we are a startup, one of a kind, especially in Silicon Valley now, in terms of we really care about the things we want to do while keeping our essence and really pushing for the maximum amount of value we can create as a small entity. And we're not doing a service to ourselves if we are not making sure we always have the people who are pushing the frontier of themselves and machine learning here.

How do you think about culture as you scale? I'm curious how you thought about burnout because people work really hard at scale and how you think about retention more broadly because there's an interesting tension where there's institutional knowledge and so you want to keep people over a long period of time. Ideally, at the same time, there's also what got you here won't get you there. and you need to uplevel and sometimes people do burn out and you need new people. So how have you thought about this?

Well, I'll talk mainly about my team so that like it's more specific and it's the things I can control, right? I think it's correct. Like we work very hard, but it's because I think we are in an industry where there's so much opportunity to be able to change the world in a better way. And I don't mean this only in the moonshot projects, right? Like everything literally I do day to day right now literally can be improved with machine learning, can bring more value to the world, not only just my life or like our lives for convenience, but can enable things that have never been done before. Think about, you know, really, I mean, improving and fixing supply chain delays, right? Improving industries that have never thought about the, transformation, improving workplace safety, improving driving safety, and not even like making our people drive, like across the board. So I think there's that urgency in all of our hearts. And if you're joining Scale, you have to have that urgency. This is not a job of tomorrow. It's today, right? We really need to take action today. So I think a lot of people already self-select with that passion and urgency. However, it is true. I think we're very dedicated. We work hard. I don't think working hard and causing burnout for the sake of doing that is is not valuable to anyone. Like there's no company I think in the world would just like want that because ideally you think of companies as businesses and they want to keep their people and the best performing people. So the way that I handled this situation, first of all. I, as a leader, have to be very good at prioritization. Not everything has to be done immediately, even though we have the urgency about the area and the things we deliver, right? So you absolutely have to be ruthless when you're prioritizing as a leader and as a team. And doesn't only have to be me, whatever area that you're assuming that leadership position in, it can be a project. It can be a team, it can be an org, it can be a department. So it starts with that. Once you understand that, you'll actually see, okay, like things will be tough, but they can be handled, right? And second, really it's about being transparent with people about the turns and the times that you're going into. For example, If I know, right, the next month or like something will be exceptionally different, first of all, I talk about everything I am doing and show to make that better. Is it hiring? Is it making sure that there are enough people? Or is it taking the duty myself? Is it coming together as a team? But at the same time, setting the expectations and talking to people, right? Not everybody can do that all the time during that time period. So we really built a close-knit team that can be open about that. Oh, I'm actually so excited and I'm willing to kind of do this versus, oh, you know what, I have this, or for some reason I'm feeling a bit tired. And we give room for those people as well to just have a calmer time. We all need that sometimes. And I think Overall, the other important part will be it's a constant give and take, right? When you're a growing company, you're constantly getting more growth, like more duties, more products to do. So the way I think about it is like, is my team supported enough to be able to deliver this? Can they speak up? What are their boundaries? We create a very safe space to talk about that, right? Like, for example, I think there's this belief generally about scale that like everybody continuously builds and it's one type of people. That's not true. I actually personally have a lot of people with families in my team, people like who have amazing hobbies outside of work that want to spend time on. We're just like very honest about our boundaries, right? Like if, if it's a time for kids for somebody at that time, we respect that and we only will talk about it and make sure Like, if that person needs to be on for something that there is absolutely no other way in the world to be able to do that. So that maybe happens like once or zero times in a given year, right? So like, it's really understanding your people, understanding people's boundaries, creating a place where they can be honest about their boundaries. And at the same time, the give from the other side is constantly thinking about according to the things you're trying to tackle, Is this enough? Do you have the right hiring velocity? Are you performing as a leader in your hiring velocity and taking on, on the work that is a good match? So I think you constantly have to think about this to ensure burnout doesn't happen and we keep the people we want to keep. And it's not even only about the organizational knowledge, right? Like when you have amazing people, You want the amazing people to stay, right? It's like even if they documented everything and organizational knowledge could be transferred, right? It's not only about that. It's just the, the thing that makes this place unique is its people. So I deeply care about doing that. That doesn't mean that like I think any startup, I'm sure, like is like this, like it is a what I'll call like a chill work style. It's definitely not that we all came here I think I can proudly say to improve ourselves, to push the limits of ourselves, and for a field that we all feel very passionate about. So everybody feels that really in their, in their cells. And it always helps, right? Like seeing, at least for me, what inspires me was to see, look around and see people who were willing to go the extra mile, right? Like that always makes me feel that camaraderie with those people myself that I want to go the extra mile. I want to go the 2 miles for those people, right? Like to achieve these amazing things from out of this smart group of people. And I would only hope, right, like in my future endeavors that my company will be that way. So I definitely have taken a lot of things and I am taking a lot of things from the culture here. That I definitely would want to implement and ensure in the exact same way possible.

Yeah. What's something you've changed your mind on as it relates to company building over the years that you've been at scale, either by seeing how it runs at scale or just learning firsthand, hey, you were doing something one way and now you need to do it another way?

I think before, this is less that, I don't know, it's not about Scale, if it worked or not. But one thing that I learned before I started Scale, I definitely came in from a place where I felt pretty much like if you can work hard and be extremely smart and dedicated, you can achieve everything, and especially the younger inspired people. I definitely came in very biased for that, right? And it really fits with my story too. You know, like before this, I worked at a bigger company and I was inspired by some people. But when I looked at generally the kind of maybe middle layer of leadership, it wasn't that inspiring for me because I haven't seen the hard work and constantly pushing themselves in the areas that they need to learn. So I came in like biased with that. But I think in the process of building this company, building my business units and teams and really pretty awesome leaders that we brought on, that definitely shifted my view, less shifted my view, but opened the aperture more. I think there are roles that you can definitely benefit from working with people who have seen more and have been through different challenges in different areas and built that experience. And in no means, I don't think it ends up being a playbook that they can just be good at and apply whatever they have done. I think it doesn't work that way. I think every company, especially if you're in an area like us, is different. But the way they might build teams and nurture people was very interesting to me and eye-opening. So I learned a lot from that. Like, I learned to respect a lot more when you find the right leaders and be open-minded to learning more beyond that, just like young and crazy energy that if you're dedicated and like you work hard, you'll get everything done. That is true, I think, still. But like, you know, you might get things done a bit faster if you have that. That definitely has been one area. And the second thing would be, I am a very engineering-biased person. So, and I think a lot of people in Silicon Valley are that, you know, if you get that, then your building is good, like you'll get it. I think pretty much I also learned throughout the experience, it really depends on the industry and the problem you're trying to solve. Many times that may not be the only thing that's enough, and you have to think about other functions that may come across at first as a non-engineering field, whether it is marketing, whether it is thinking about building structure on the sales side, things like that. I think it definitely— I got a chance to learn, depending on the idea and the industry you're going in, those matter and you should invest. You should invest early and in the right way to be able to get the best results. So, it's not— while engineering is key and central, like, there are industries that it may not be your only path. The path to success does not only include that. It has to be supported by other things that I previously saw as, oh, I don't know if it's that necessary in the beginning in my endeavors.

Before we get onto your remaining time at Scale, I want to talk about the other side of the chief of staff equation, which is from the CEO side. One, talk about the pitch and how CEOs should be able to pitch really talented people on becoming chief of staff, what's important to get right there. And then 2, talk about what kind of archetype to look for. How did Alex identify that you would be a great fit? He could have potentially taken someone with much more experience, but he took a bet on potential. And talk about maybe the trade-offs there and how you think about the selection and the pitch.

Yeah, yeah. Well, I'm sure he would be the best person to answer this, but these are all only my—

Don't be modest.

Exactly. No, no, no. These are only like what I may think. And I'm sure there might be things that obviously I haven't seen or he hasn't shared with me. But first of all, the pitch, right? The way it landed, first of all, I think he didn't mean that this, and I didn't mean this, this happened serendipitously. But in those first few meetings, we really built a bond. And it's so funny, like, we didn't build the bond, like, from friendliness, but I just felt, I think, the room and space. And this is a little bit about my personality too. Some of the ideas he was talking to me about and, and running through me about, like, why we have met, the, the certain applications of, uh, these products in different industries, I felt The comfort level and the confidence to just say they're not good enough, and and I was even meaner than that. I think like I might have at some point said, "Oh, these suck." Like you should not do that. But I think that a little bit kind of took away the armor, right? In the sense that oh, like interesting. First of all, I think came across to them. Wow, this this person. Like I think as CEO, when you're CEO, not a lot of people like feel comfortable saying that. you are bad at something and you haven't done a good job. I think that really was interesting, right, for him. And from my perspective, I don't know if I would have done the same thing, right, if we met from a recruiting perspective. Obviously, when you're going for a job, you want to be liked, but I felt that space. And definitely it's a personality thing too. I'm just a very blunt person and I feel like things should be said. And if it's helpful, And I have strong opinions that are loosely held. If you show me different data, I might change my mind, but I do have the strong opinions. So I think that really helped. He was like, oh, interesting. Like, this person is trying to tell me the truth and in a way that is very effective and direct. So that was definitely one of the things. And second, I suggested some ideas on the call as well, right? And they were very quick, like they were on the go. I didn't even know the problem. I didn't prep. It was just like my style of brainstorming. So that shows something about the capability and promise, right? And I believe we both did afterwards, like after the initial interest. So when he came to me, I'm sure before he pitched me the role, he did a bunch of references. And I know that he talked to people who know me. And I think he really heard the things he was looking for. And there are a few things Alex deeply cares about, like, is this person really smart and creative in a way that can create outcomes autonomously? 2, can I trust this person? And 3, is this person really hardworking? They're willing to really give it all they got? Because in an early startup, you can't afford the other way. And I think he heard things that resonated and checked the boxes in these areas before he called me for the pitch. And in the pitch, I think first of all, he really understood well what I wanted to do, like at that moment. And if he could get me like afterwards, like he knew that I'm going to be fixated on that goal. And, and so it's important, like to understand what the person's real desire is. And I think a lot of people don't go beyond like how amazing they are. But if the person's real desire is different, it doesn't matter how amazing you are or how you're pitching. Your own company or yourself, right? It has to be aligned with what the person is. And I really care about this today in my hiring. I really want to get to know that person, their goals, their motivations, and make my best to be able to meet those motivations and give opportunities for those motivations. He did that really well. So the whole thing was around, right, the kind of things that I can do here that will— first of all, that I can't do anywhere else. Like whether it's startup or a company, right? Like I think there's this thing that like many times in these companies we don't allow the people who are capable and young to go ahead and run with it, right? And if they can deliver, they get more and they deliver more. So that was like number one. Number two, that how I would get a chance to work on things that I believe will make me a better founder. And three, I think it was that personal connection that made me feel that he also pitched, that I would get a sense, right, how he's making decisions, how he's figuring it out on the go, and how I can figure it out. And he would be that partner, like genuine partner in crime. And I think if you're ever gonna do a chief of staff role, like, this was basically perfect according to my goals, right? If you don't have the right goals, this job might be miserable for you because, like, when I worked on— building things from zero, literally at 3:00 AM, like I was thriving. I literally went to bed with like, oh my gosh, like I need to wake up and do this again. Right. If you're not looking for that, it will be miserable. So it's very important to understand if it's the right fit. And yeah, that, that was definitely the thing that resonated with me. And the second thing was his responses when I went through that granular list and really held him accountable for the way that he was going to talk to me. He was very honest. You know, even in the questions I asked, like, he didn't try to sell a rosy picture. He was honest with me because I think he knew that, one, I can find out, and two, like, I really value honesty in my life. If I don't see honesty, I don't want to work with the person or at the place. And yeah, I think he was very mature and, and understanding why I cared about one thing versus not the other and how it would meet his needs. That negotiation exercise, and I mean negotiation in a more values and the job description perspective, not the numeric negotiation, but I loved it. He was very himself and I was like, this is a person I can really be myself around. That really went well for me. And I think he didn't really, After the pitch, I also didn't obviously just trust this. I did my own homework and references. You know, I talked to literally everybody and it's very important. I have one rule when I, uh, before I decide to make a hire myself or go into a place, I don't take only positive things. Everybody in their lives about who they are has at least one terrible negative thing about themselves. I have this. Right? Everybody has it. So if you don't know it, that means you haven't looked enough. So I will not stop in any case if I'm trying to learn something, if I'm a good match with somebody, until I find something I'm like, hmm, interesting. But knowing this, I still accept and love this person. So I definitely did that too. That went well. And I think even then he still didn't like— it's not like he was like, okay, this is capable, here is everything, didn't trust me. It was the first 6 months actually that made the decision for him. I'm pretty sure when I first started, he thought there was still a huge risk I wouldn't work hard enough, probably because I was coming from a very big company where there's a general conception of people like who are just leaving very early and showing up at noon. I think he just saw it with the first thing that I undertook with building the government business unit. He just saw me going at it like every day, every hour, every night. And even when I got nos, I didn't give up and I came with a different creative idea. So I think it was, I would truly say it was maybe the end of like month 3 that really made it for us both ways.

Totally. Let's spend the remainder of the podcast talking about your role after Chief of Staff. You built up and scaled a couple business units. Why don't you talk about what you did there, how you approached it, and what you learned from it? Yeah.

One, actually, the first one is while I was still Chief of Staff and working on many different projects, sometimes I even like kind of took the role of different functional leads, like in marketing or in sales. Like, I would do the sales myself. But I was very interested in the government industry. And there was a reason for that. I have done some work in college at Stanford with Hacking for Defense and working with different government agencies. And obviously, there's an increasing pressure also when it comes to machine learning and how we are in the world arena, right? Like how much it matters for the United States to be knowledgeable and understanding and build the right technologies when it comes to machine learning. So I was very interested in that, and that was a good opportunity. So I kind of noticed this opening and started taking maybe a few conversations. In the beginning, it was like not really much because it was just like researching So hit up some people I knew from the past, but also I just found people like literally on LinkedIn and I was like, this is the right person to talk about machine learning in this area. And they would take my call. Like you would just say, hey, I'm working here, I'm interested in this. And here's like a common person we know that said like, you'd be great. Can you please talk with me? Like have coffee. And I remember I would really try to meet them in person. So like I would say, oh, how about like within 2 days, you know, I'd be like, in 2 days. And if they said yes, like, I would fly that night to DC. Like, basically, like, you kind of set things up and then, like, fly, go take those meetings. Through that, we learned a lot about the field, like what they need, how we can help, how we can shape this in a way that we would be proud, right? Like, as people who are really committed to machine learning and the people getting use out of it. Yeah. And then was very interested, started with these knocking doors, to be honest. And like, he would join me as well. Like, we would just go to D.C. a lot and really try to understand problems, not even the solutions yet. Like, listen to people, build that relationship, and build different demos and products to show them what it can be that really resonated with that audience. I think like we many times have in Silicon Valley, thoughts and criticisms and feedback for the government industry. But you know what, like, we're also not always like making it easy, right? Like, it's very difficult. We're not listening that much. We're not understanding the problems. We're not building that relationship. I think doing that really helped. And then from there, I got a small opportunity landed. So it became all about execution for me because in the government industry, it really depends on your execution and your reputation and execution. execution to continue. So really gave all I got in being able to deliver absolutely the best thing for this opportunity we have gotten. And we did, right? And use that to build a bigger business in terms of getting bigger contracts, building those contracts, writing those contracts, thinking about the delivery of those contracts from an engineering perspective and getting to know more problems across the way. Really use that small opportunity to be able to do that. And then as like the next step, basically this big contract that came up, the first one, the big one that we have gotten around like a $90 million business. So like really worked hard to write that and like learn how to write it, like figure it out on the go and write really the things that we can provide and build an amazing team that can deliver on these. And now it's really our garment team is growing. Really fast and solving problems that are so incredibly important. When I think people think about ML and government, there's only very select use cases they can think of. But today we take so much part in disaster relief, right? We take so much part in like what are systems that we can build that make it more efficient for whether it is the supply chain or logistics or disaster relief or, you know, like helping war not become war right now. Right. Like, it's so relevant right now, like understanding where are the danger areas, how can we get people out there, like things like that. So that really made me passionate. But at some point, as I was also rolling off of chief of staff, that was my one kind of condition. I didn't want to do it for a long time because I want to keep on my own personal goals. I did not want to continue in the government side because I know that my own future endeavor will not be in the government area. So I wanted to look back in the commercial and found an opportunity, saw an opening about the thing that I do right now, which was like, if we're so good at building the data infrastructure layer, it enables for Skillz to be a great player in the stacks on top of the data infrastructure, right? Like on top of getting the high-quality training datasets. And one of those is models, actually building the machine learning models where you need high-quality datasets and you need the best ML engineers in the world, which we have both. So I saw this great opportunity and wrote a plan, actually, like literally wrote a plan about why we should invest in this. And that I want to do it, and here is the you know two people I need to start with, and presented that plan and kind of got the the go. And I think it's a very entrepreneurial environment here. If you see an opportunity and you can prove why it's a good opportunity and that you can deliver, you will have a chance to do it. And have been building that team since then. We started with maybe. To people trying to do everything. Now, even with the extended team and all the functions across the board is like 30 and counting more. And I lead an awesome team full of great engineers, machine learning engineers, sales, marketing across the board, really to bring machine learning to a real-world problem. Which is like, how can we actually be in 2022 now and process, understand text so we don't need human reliance in a lot of the key industries that affect our lives today, including logistics and financial services and healthcare? So have been focusing on building the best models on that since. So the role shifted to become all about one area. Before it was right, like, It's whatever life throws at you, you have to be good at it. Managing 1,000 things at once with the high quality. It's still 1,000 things at once, but all of them are related to one thing, right? Before, it could be, I don't know, people and engineering and sales and then infrastructure or a contract, et cetera. Now it's all around this, but still variance in terms of Building the right team, building the right product, getting the right feedback, going out there, actually being the person selling. I really care about a lot about that for new products. You have to be the one hearing the feedback to immediately continue building the right product. And the best part, the most challenging part, is also the hiring when you have your own teams and building it. That that's. That's the part I have learned a lot and it's been going amazing. And I always really make sure a significant chunk of my part goes into bringing the best people as we grow and really making sure that this team represents the right things and the right dedication.

In closing, talk about how the business units were situated within Scale. and how you thought about org structure and what you sort of learned about org design at scale?

Yeah, yeah, that's actually a very good question because, so there was not really any business units until the government one that I was building actually at scale. It mostly, the way that it worked was we have a big engineering organization that covers engineering product and design. And then there's an operations org, there's a go-to-market sales org that encompasses sales and marketing. And then of course, like G&A and people functions, right? So the way that the products were built, there would be an engineering team for that product. So let's say like on our Labeling Pro products, let's say it's a data type, 2D images and videos. So there would be a team of engineers for that team and an engineering manager and a product manager assigned, and they work with their counterparts on the sales side depending on the industries they're going after in the go-to-market and really have that partnership. So the business unit structure was new. It was very natural for government. Like, we didn't really have to, quite honestly, think a lot about it because it made sense. It's new. The requirements are usually different. The level of infrastructure you want, you need to do different. And there might be even sometimes like restrictions that you have to keep it Separate. So we started with that, but that was a good case study because when we did it that way, for something that is meaningfully different than the rest of the business, and we need to be fast, determined, that structured work worked very well. It really opened the path, right? Being able to just very fast hire people to that unit, not really creating bureaucracy around these like various approvals, even in different functions, having a strategy that is close-knit really accelerated us. And I mean this. So when I saw the other opportunity on models as a service business units in building these models, I wanted to replicate it. So basically in my pitch, not only was about the product and the path, but I added why it should be in this way in terms of the org structure. And my hypothesis was one, in terms of the distribution channel and the industries we're going after, right, in this area, it's different than the things we have done before. So you actually don't want it to be— you want to take the best lessons that you learned from the other side, but you don't want the new of this be affected and shut down by the learnings or the things that you were accustomed to from the other side, right? You want that open-mindedness. And 2, obviously we're trying to bring to the industry a better machine learning model, and the industry was not even there yet in terms of understanding machine learning itself. There's a lot of education, the things you need to do very fast without being blocked or having to, you know, go through a lot of steps in the rest of the company, even though they're still small for us, like as a startup. But even then, you don't want that. You want to be very, very fast. So that was like really the second argument. And the third part was the interesting part was having the sales underneath, right? Is that it's actually in building new products, it is so important to be there to understand people's reactions. What did they see? How did they react? Did they want something else? Be that technical and product leader to even able to walk them through about how we might be able to solve their problem or how can we even assist with their problem statement was key. And I was saying basically that this is very important for us. This is a very new product. And for us to be fast and build the right product-market fit, like, we have to try it this way. I need the acceleration. And then the final thing was, I think in org structures and organizational ideas in the company, you can't treat them as your baby. Like, you can't be attached to them personally. You have to be mature enough to let go if it doesn't work. Okay. So my last thing was basically in that proposal was, you know what, like, this is a great experiment because if it works, we can replicate more. But if it doesn't, by all means, we can also just end it, right? It's very easy to just like split it apart and like make sure they go to the functions as we have been already doing. I think that gave the right feeling about like my motives, the business unit's motives, and the great experiment because we always continuously innovate in our org structure as we grow to empower smart people and driven people, entrepreneurial people. So, it was like it had that attractiveness, right, that we can try this. And if it works, it's actually a good idea to empower more people to build more products. So, so far it's going well, but the jury is still obviously out there. We'll probably it would be a great time to evaluate in about 6 months. And for me, the end goal is always when a product becomes mature enough to fold in and follow the organizational structure that is working, but bring new products that can have this business unit structure, right? So you kind of have the, your, the acceleration part, like new product incubation part constantly being super fast developing And as they find product-market fit, can come in and be part of the mature products as you continue the evaluation for your business units. But we'll see. I'll let you know in 6 months.

Yeah, that's a perfect place to wrap. Melissa, this has been a really great episode. For people who want to follow your journey and learn more, where might you point them?

Please, you can follow me on Twitter. It's @MelissaTokmak, just my name and last name. And also on LinkedIn, I'm happy to Talk to folks, especially if you're really excited about machine learning.

Awesome. Thanks so much for coming on the podcast, Melissa. It's been a great episode.

Thank you. Thank you, Eric.

Thanks for listening to today's episode. Execs is produced by On Deck, where top talent goes to start companies. find their next roles, or invest in their careers. If you're looking to start a company, uplevel your career, or navigate a career transition, I encourage you to visit beyonddeck.com. See you next time.