What Senior Engineers do Differently (Vercel VP)

It doesn't actually matter what your level is, you could impact the business if you figure out the thing, if you spend the time. This is Lindsay Simon, VP of Engineering at Rocell, and he shares what great software engineering looks like today. Have you felt FOMO? The last few years. Yeah, I felt FOMO my whole life. The people who will lead and come up with the innovation and to be the people building. They're really good at unblocking themselves, and they're really good at unblocking other people proactively. We're all junior engineers again. Have you felt FOMO in the last few months or the last few years? I think that part of that is the drive to prove yourself wrong. I have worked in a variety of different environments, tech and not tech, where I felt like the challenge was proving to myself by proving to the people around me that I could do something useful, productive, valuable. Because in my head I'm thinking, I probably can't do that. I can't hang in that environment. The only way to find out is to try to do so. If you get a hook in for trying, then all of a sudden now you've got in, but now it's even worse. Like, shit, I got lucky. In 2006, I interviewed at Google, and I had no expectations. I had no reason to believe I belonged or could get a job there. But I did. For whatever reason that day I got lucky. But that wasn't the end of it. Now I'm among this group where I feel very incompetent. I'm like, what do you do? You can freeze in that moment. You can struggle. But you got to fight. It can heal you. That's how I've been my whole life. That's really cool. Looking at our industry, I think agents are shaking up a lot of roles and responsibilities. Do you foresee any roles just disappearing in the next few years? I mean, I think I see them consolidating. There are roles that were created as independent roles. I mean, they weren't ordained from on high. There is the engineering manager. And these are its jobs. And there is the product manager. And these are what they do. I think there are companies and times when we try to optimize by building lanes. The idea of you could do two things at once. Well, you had one person doing this job, one person doing this job. And now you're going further together. I think engines definitely change that because now you can say, especially finding out where you are less well trained. It's a great place to engage with agents because they have access to the world of information that you probably haven't read. Or spend as much time learning about. And that's a great way to build in that area. So I think engineers, the trend in the valley right now is that everyone is becoming a member of the technical staff. Yeah, I was thinking the same. Which tells you that I think there are two reasons for that. One is an anti-poaching vibe. You don't know what this person does. We're not going to tell you publicly. I never thought about that. But internally, they might be a TPM. They might be an EM. But externally, they're just a member of the technical staff. And I think it's twofold. I think it's the anti-poaching piece. But also, you know, there is a flattening of an sub-semation of like responsibilities and ideas. You know, you might not be a designer by trade or training. But that doesn't mean you cannot do design. The active design is, you know, multiple things. Have you seen that shift also then in not just from engineers and the capabilities there, but also in hierarchy? Yeah, definitely. I mean, you know, I'm historically, I think I was nervous about, say, new grads. You know, less experienced engineers joining the company when onboarding was really hard. But, you know, we have interns now who show up at our demo days, which is a Friday. You know, anyone can sign up and just get in front of the entirety of the engineering team, including the founder, the CTO, the engineering team. You know, you could be nervous or you could be excited. And you're all equal in that environment. And I think that's actually one of the things that draws people to tech is the egalitarian. I mean, the meritocracy idea that it doesn't actually matter what your level is. You could impact the business, you know, equally as other people. If you figure out the thing, if you spend the time, I think that's a very attractive part of technology that will not go away. For those people, then specifically early in career that you've seen Excel, what makes them really excel in what they do? Because I also see this narrative of, oh no, what is the junior people? Yeah, I mean, look, I'm perfectly happy of the rest of the hiring community thinks, oh no, what is the junior people? I'd be happy to work with them. What I see is that, you know, they're growing up in a builder-minded culture. Where, I mean, once you have access to these tools, you're not limited by, oh, I don't know back-end engineering. Well, I mean, you don't, but you have skills and, you know, you can matrix your way, you know, eat the pill, learn kung fu. All of a sudden you have access to knowledge that you don't directly possess, but which you might be able to leverage. And so I think there's less hesitancy for people that, you know, don't have the burden of experience and thinking you need the certain kind of expertise to even pretend to do a thing. You can be a lot more bold. Well, I don't know anything, so I may as well do everything. I think that can be pretty powerful. So I see, you know, college students who've been doing three years of hackathons, doing internships at big companies and in tiny startups. And I mean, they're coming out of school very, very well prepared to go work in with other people. You know, which is quite different than say you went to do, you know, very hardcore academics for four years and then you had two jobs, you know, three years each. Yes, you have six years of experience, but that new grad might have four years of experience and it might be more relevant. So I think it's an interesting, it's an interesting dynamic. Yeah, one of the pieces of advice I got early on in career, it was always to reach out when you're stuck and collaborate. Because especially in traditional education, you are kind of taught to figure it out yourself, right? Dive deep when you can learn because learning by doing actually makes it the still your knowledge and it anchors it better. But in a business sense, if you're really stuck and you're going to be stuck for weeks, just ask a question if that really helps out. I don't know nowadays what the impact has been on collaboration for people that work with agents because I can ask you or can ask my agent. And it's easier to ask my agent because then it doesn't really look like I don't know. I'm trying to figure it out myself. No, and I think that's part of it. I mean, when we had our first cohort of interns in the post GBT era, they onboarded very effectively because they were asking the kinds of questions that, you know, probably weren't the ones you'd ask somebody. You didn't need somebody who had four years of deep experience in the code base to answer them. So you could ask, you know, GBT. But in, you know, and now you're going to figure out, okay, you get good at it and looking at the responses and going, ah, that one, that response is kind of fuzzy or it looks generic. I think I will take this question to somebody who's gotten more direct experience. And now that experience is grounded in, you know, you've thought about the problem even a little bit more. You're not just asking because you're stuck. You're asking because you tried to unblock yourself and you're not satisfied with the answer. That's actually a great point to, you know, and also the people on the other end of that will appreciate it. You know, hey, you know, you didn't just start by being lazy and asking me. You did your research. You're unsatisfied with the answer. And here's why. That's a much better way to frame asking for help. Yeah, that part of being unsatisfied with the answer, that for me is very crucial because you could take an answer and be like, well, this is the truth. This is good enough. Yeah. Or you think and you're like, well, this doesn't really make sense based on what I know or based on conversations about that. Among the skills that I think we get better at is really questioning, you know, you know, it's not enough to go, oh, well, this is the answer. I was like, well, why is that the answer? Okay, and you keep kind of going backwards, right? You're digging backwards into really a fundamental understanding. And if you build that fundamental understanding back, you actually really can understand the problem space you're in and be much more effective. Yeah. I draw the similarity to search because previously if you find a source, it's like, well, where does the source come from, right? Is it a news tabloid or is it actually true for information? Yeah, exactly. My ten year old daughter was doing a class on how to evaluate good source information from search. And I think they're actually teaching this. I mean, that was not a class when I was in school, obviously. But I think that kind of education going back into elementary school tells you that somebody coming out as a new grad probably had some of that kind of experience very early on. So when you say, oh, they have four years of college experience, their education experience is entirely different, you know, being internet native or even AI native, right? Yeah. And that's what we mean. That's what we mean. It's incredibly crucial that you still are kind of, you can trust the results, but also just keep thinking and be skeptical as well. Yeah. And then validate with your peers because that's what they're there for. Yeah, I mean, that's the skill. Learning, you know, the learning of when to when to push back, when something doesn't feel right. And you only know that by really building an understanding to question things. I was a bit for you because I recently spoke to it was the city of Nebius. They are an AI cloud company actually quite on the forefront. And that person said, well, I'm a manager, right? I'm a leadership team, but I can now build as well. Yeah. And he's super enthusiastic about that. Yeah, I'm enthusiastic too. Yeah. So now building more and more again. Yeah. Yeah. How's it been? It's amazing. I mean, it's very, you know, we're all junior engineers again. You know, these tools haven't been out for longer than any of us have been in the industry. So it's, you know, that FOMO that we talked about earlier. It's back, but you know, so what do you do with it? Do you kind of hide and you know, you scared, there's the tendency to do that, but then the only way out is through, right? You have to really come back to it. And I think that's what got all of us into building in the first place is the art of crafting and building and making things we'd want to use that we want to share with other people. And so, I mean, that inclination hasn't gone away. I mean, more motivated to do it now. It's exciting. Absolutely. Have you seen a shift in how people want to traverse their career? For me, it was always you go deep in individual contribution or you kind of go up in the hierarchy, engineering manager, director, CTO, those types of roles? I mean, I think that, I mean, one of the things that we tried to do very concertedly early on is to not equate management with leadership. I find that, you know, that's sort of the default language and taxonomy, but it's not what I'd want as a, you know, my heart, we didn't, I didn't never start, it's going, you know what, I want to be a manager. Yeah, I want to manage people. Yeah, I mean, I don't want to be managed myself. I'm an unmanageable, you know, you don't really want that. You would like to have peers and people with more experience who can guide you and help you find areas you should, you know, see the things you might not see about yourself. I think that's great. But, you know, and for us, the IC track and the management track are relatively parallel. The highest level on the IC path is as high as the highest level on the management path. So, you know, that it isn't like you pick one and there's a career ceiling as a consequence. And I mean, I don't, I don't think that's super novel, but I think it is, we are going to see this breakdown more. And that's good. You know, we didn't, the people who will lead and come up with the innovation at the companies that I've worked at tend to be the people building connected to the iteration cycle, you know, talking to customers, you know, you don't let leave that to somebody else. If you're doing all of the jobs, you really can own and excel at something and make it succeed. - Yeah. With now the ingredient that is agents, for me, if you asked me a couple years ago, I would have said, well, engineering management, like one of those career paths was definitely something I saw for myself. - Yeah. - But now I got the biggest phomo because if I'm not hands on, I feel like I take myself out of being a daily practitioner and individual contributor and things are moving so fast. - Yeah. I feel like I need to stay here a little bit longer, especially since we're accelerating more and more. - Yeah, I mean, there's a push in the pull. I think knowing I didn't end up becoming a manager, except to challenge myself to see if I could help, you know, scale impact outside of my own contribution ability. For most of my career, my, you know, limit was how much PR through what I could achieve, you know, and add value that way. So what if you take away your, you know, number one lever? Like that was my lever. And if you take it away, now what can you do? I see that as a different challenge. So in that way, not having that and seeing what else you can do, it's just another challenge. - Where do you get your fulfillment out of? Because PR's features, users actually using what you build, is very different than, well, now you're in a central role, kind of orchestrating, but people are doing really good and then is that you or is that those people? How do you get your fulfillment? - I mean, I think in the end, the fulfillment of the mission of a company is usually, it's what always drives me to get attached in the first place. And if I can continue to feel that, you know, it's not like a multi-day adrenaline rush of landing PRs. It is a little different. It's a, you know, spaced out longer sometimes, but sometimes it feels, you know, when five people all feel proud of their work and excited and you've helped them do it, I mean, that feeling of enablement, that's part of the mission. So, you know, even Versel as a company, its mission is an enablement mission, enable developers to ship the best products. And so when I see a company building something on the platform that, you know, that's a great idea. I don't think, oh, damn, I wish I'd thought of it. I think, wow, they're building, we help them do this. And this is incredible. And, you know, you feel pride through that. It is, you know, maybe that's a function of age to some extent. You know, like, you know, I wouldn't mind getting older and becoming a teacher to help people learn how to learn based on what I've experienced. And I think maybe that fulfillment is a little different to different kind of richness. Yeah. It's, you know, it's group dynamics. So I get it. I feel it. How has it been building specifically for the audience that is engineers? I feel like engineers are quite specific with their taste, what they like. If it's shit, they'll tell you definitely. But also in the last few years, that taste might have changed. Yeah, I mean, it's a conveniently engineers are a fickle lot. You know, and there's not a single answer. In aggregate, you do get, you get a lot of signal. There's the signal you get in social platforms. And then there's the signals you get out of telemetry. You know, not everybody's spending all their day on X telling you their opinions, but they vote with their feet and they vote with what saves them time. And so that telemetry is extremely valuable. And it's the thing that I'm pretty confident as we continue to push on improving telemetry, you're actually really changing user experience in this super fundamental way. So those two things are both good. You know, I value social evidence. And you know, we read a lot of feedback, harsh and positive. And that incorporates in shapes the feeling of, well, what can we do with that? Could we, you know, somebody had a, you know, there's the feedback, you're like, oh man, they're so wrong. No, no, we didn't do a good job. If they, they're not wrong. Nobody's wrong here. If we think they're wrong, then we didn't communicate it well. Yeah. You know, there's something broken that we need to fix. And it's similar to, you know, I used to do a lot of user interface development. And if you sit somebody down to blindly try your web application or your mobile app, and you know, they have a little less context than you do. Obviously. A little. Yeah. Just a little. And now you watch them try to use it. It's such an incredible experience. I mean, usually it is the most motivating thing because you're like, oh my god, you're seeing it through somebody else's fresh eyes and you're seeing them frustrated and lose time. And it's exact opposite of what you'd hoped would happen. Yeah. And now you're like motivated. And I was like, oh, I could fix this. And then, you know, you're like, stop it. Don't show five people. Let's go fix it first. Yeah. Because usually they would line up like five interviews. And you're like, no, I have to fix this now. And that urgency from feedback is that's the dream. I love the mindset that you say, well, if there is a signal, right, don't just blame, don't just put them in a box, but actually take the feedback and take it to heart and be like, what did this come from? Analyze it, look for its value. Face the friction. Yeah. It's friction. It came, that energy came from somewhere. And it's there. Yeah. And one of the most fun experiences you already mentioned it, having people actually go through your application. And then not being allowed to say anything or steer, it's hilarious. It's really funny. It's knuckle, you know, knuckle whitening. You're sitting there in the other end, just dying. Yeah. Have you done that in some of the products right now at Rosalio as well? Oh, yeah. I mean, we share feedback with a lot of, you know, we'll do just very impromptu feedback sessions. And it's educational. And I think we take a lot of pride in also providing that feedback back to other companies building on the platform. You know, we work with-- I mean, the slide deck I presented here today with was through a partner company that's built on Rosalio and they built a slides product. And so I sent them a ton of feedback about my experience building my slide deck. And like, you know, that sort of back and forth-- and I expect them to give me feedback on the AI primitives on the platform that they're building on. And so, you know, you've given you take in this. And that's actually-- that's pretty magical. Yeah, definitely. I want to shift gears a little bit and talk about expectations. Yeah. Because I don't know if, from a leadership perspective, expectations for senior staff or people that want to level up in their career on the engineering track have changed. We already talked about kind of being skeptical of the output, trusting your sources or validating with peers. How have expectations evolved in the last couple of years? I'm not sure they've changed super meaningfully. I mean, perhaps the baseline for certain things has come up. But I don't think the general expectations are different. I Google at this idea baked into the performance ladder of scope of impact. And I've often thought about leveling up in your career as a function of increasing your scope of impact. You have yourself as a really competent IC. There's kind of two levels. There's starting and experienced. And then you have the version where, you know, you're a tech lead. You're actually helping delegate and drive work for other people. And you might be new to that. Or you might have really started to master that skill. And then there's the version where you're doing that for multiple teams, kind of consulting as an Uber tech lead or as a multi-team tech lead. And there's newer to that role. And then there's very experienced at it. And then there's doing that at the entire company level. And then there's doing it at the industry level. And I actually think that maps pretty well to most of the levels in most career guidelines. So I'm not sure it's changed in that too much there. Success might look different in terms of speed expectations and things like that, because we do know that the act of coding can be done much more quickly with agents so that may be different. >> That's what I thought so. >> Something that you can, Cloud tells you it'll take a month and it takes two hours, maybe one. So it's even bad in its own expectations management of time. So I think that is not such a, that example comes from, well it's trained on probably JIRA and the flow of tickets that it has access to and what it expects it would take. And then in reality, it takes a 50th of that time. Well, I think the expectation is some things might take a 50th of that time. Not all things. If I were to say, think of the best engineers that are kind of in your work environment. There's probably a few names that pop out. >> Sure. >> What do they do that you really appreciate about them in their way of working with the impact they bring to the table? >> Yeah, they unblock themselves. They're really good at unblocking themselves and they're really good at unblocking other people. Proactively. You know, they're in touch with the kinds of communications that are happening in the company. You know, they're not, so they are, they're reading and they're writing, but they're helping. >> Yeah. >> And then that help usually causes them to systematize their help. So, you know, that's a way to scale how you impact. If you figure out how to systematize, helping two people directly and you leave a breadcrumb behind that'll help the next 10 people, you know, that's a serious, that's the kind of activity that happens with at a more experienced levels. I see that. Do you have an example of what that system occasion looks like? >> Yeah, I mean, an easy one is that you might, you know, have years of experience building interfaces and, you know, shooting, who works at Versailles. He wrote one of the first skills that we started to think about. How should we distribute this more broadly? It was his like best practices for react applications, built from years and years of experience, you know, debugging and fixing async problems with fetches. You know, instant mutations in clients and on servers. And like, codifying that into a skill, and then us distributing it, we've given you shooting. >> Yeah. >> You know, and shooting has, you know, changed his impact scope way beyond, you know, industry level. And I think things like that, you know, the ecosystem for sharing reusable skills and plugins, I think is a good example there. >> Yeah, that's a beautiful example. >> Yeah. >> Yeah, definitely. I feel like education is also kind of shifting. And people definitely when new things come out and the market, you also see it, but from a marketing side, but also from a content creation side. Right now we're at a conference, TechLid conference. >> Yeah. >> What do you appreciate about conference environments? And then this one specifically. >> Yeah, I mean, the conversations are a higher fidelity than anything, and, you know, people open up more personally in person for sure. >> Yeah, you're kind of more aware of your blast radius. And therefore, you kind of open up in a different way. You know, seeing people at very different stages of their experience level. You know, some people very hungry and a little scared. Some people very experienced and scared, based on the state of things. But usually they're here at all. It means they want to learn more. This tends to foster an environment of self-learners and people trying to, you know, get better. For the sake of humanity, I think. You know, that's pretty inspiring. >> Absolutely. Have you seen any pattern in kind of the appetite for topics that people are talking about or asking about? >> I mean, the one that's pretty fun that I noticed on our panel. I mean, my talk was the first one of the day. So I hadn't seen the other ones. But the idea of people going back to the fundamentals of how to learn effectively, I think is really exciting because it invites us to go back to sort of the stages, you know, for me. You know, I've been in industry for a while. And, you know, going back to, okay, that's built on a education where I learned how to learn, learned how to write by doing it over and over again, and critiquing the hell out of it. And we still do that in the industry. But, you know, what can you take as principles that you can apply to this time? From something you might have done much earlier in your life. And I think, you know, finding those gaps where, okay, well, I got through this class and I got a D. You know, it's clearly not as good as you could have done. Or maybe you got an A, but along the way, you know, you passed something, you know, you got an 85 on a test. You know, I think the model today would be, go back. And not just today, this is always the case for what effective thinking meant. Go back and really get yourself to 100% understanding and then move forward. And you won't have these gaps that sort of grow and create splinters and a different kind of FOMO that is legit. - Yeah. - You know, if you close those gaps, you have this opportunity to figure out for yourself how to build on stronger foundations. And that helps you to think more clearly more, and then move more quickly. - Yeah, I really had to learn that by doing, because I kind of moved into a software engineering role and then I learned to be effective. But then I wasn't as fast as my peers or things just took longer than I initially had thought. And that was the underlying foundations that I still had to catch up on or kind of refresh. And because things are accelerating now, I love this narrative or this signal of, let's learn how to learn from a fundamental standpoint. - Yeah, exactly. I mean, I think, you know, there was always this concept of, well, should you drop out of school and just start a company, should you, do you have to study computer science for four years to even get a job in the industry? I think there should be a resurgence of liberal arts education for everybody. If English is the most important programming language, you better learn how to write and write effectively. You know, distilling information is a practice art. And then communicating concisely and, you know, with measurements and, you know, I think a lot of this applies to using AI effectively. Because you can burn a lot of tokens to get there along the way. And again, that's more slow. And that isn't what you're trying to do. - No, the distilling of information, I never thought about it, but it's something that we always did with all the various sources of content that we have now. - Yeah. - But especially if we go to more leveraging one source that is AI models agents. - The context is huge. And then it's up to you to condense it and be like, this is the essence and communicate that. - This is the essence, yeah. - Yeah, yeah. - I think that's one of the attractive things to people in this, in our industries, that symbolic logic is it can be a very concise way to represent pattern matching. I mean, I was fascinated by regular expressions early on because they're so powerful and so few characters. Now, you know, you could argue that they're not very readable. But in fact, if you learn how to read it, it's very readable. So I think that a lot of things kind of fall from that. - Yeah. Before we round off, I want you to share one takeaway that has helped you within your career that might be relevant for the people trying to level up. - Yeah. Yeah, I think this comes back to feedback. Don't be afraid to get feedback really early and continuously, continue to do it. You know, you don't wanna go into the cave, you know, use a stacratic analogy. You don't wanna disappear into the cave. Digging, you're gonna come out with, you know, an earth-changing thing. The way that all great practitioners have developed is by constantly getting feedback from more people and learning how to distill that signal. So, you know, get out there, share your things. I mean, for me, writing was the big, most important piece of advice I ever got. Was start to learn how to write about the things you're doing. 'Cause that forces you to turn, you know, experiments into concise descriptions. And that's actually very interesting to other people. And so, you know, get out there. - Awesome, thanks so much for coming on. This is great. - Yeah, thank you. - Thank you. - Thank you.