Sergio Prada, Metal: YC, SF, NY and AI

guys welcome to another founder wisdom pod we have Sergio Prada with us today he's co-founder at metal uh part of the YC winter 23 batch uh interesting startup to say the least me and Sergio were free speaking to that call and I was just like quite impressed with Sergio's background Sergio can you introduce yourself tell us a bit more about your eclectic and interesting background as an engineer oh yeah absolutely thank you so much Charles uh I'm Sergio um I I've been working in software for the past 10 years or so um I'm from Colombia specifically uh I studied there chemical engineering I started some startups there I actually dropped out to start a startup uh which it failed miserably uh but it taught me a lot of things and I kept working I started working as a as a consultant after that and then building websites for uh advertisement companies and marketing campaigns then I kid building and building eventually I moved to San Francisco where I was in a hacker house for like six months or so um building just more knowledge and portfolio until I was finally hired in New York where I've been living for the past seven years um and there I I've worked in multiple companies big in in small ones I was the second back-end engineer at a startup called customer which got acquired by meta and then I was there at meta for a year or so but um during my time in New York I've been also part of datadog um Shutterstock and so on and now I I am the founder of metal a co-founder of metal uh we've been working on metal for about like four months now uh we just finished the batch winter 2023 interesting so let's make a deep dive in your past like did the Young Sergio in Colombia ever pictured that the future Sergio would be a software engineer uh making it big uh at like big companies and also starting a startup like how did your childhood and your past influence your your current self I.E yeah I think I've never I've always had this dream of starting my own company um but the dream of going to YC and building my own startup working here in New York City I think that's more of a recent thing uh I've been coding since I was in high school I actually started building websites for like classic gaming Clans um I remember student building websites in geocities and um very very uh a long time ago but I yeah I've always had like a clear idea that I want to start something uh but it wasn't clear until I was here in New York City that I wanted to do the startup in in technology and that I wanted to um build a company here in Brooklyn and that's that's so far so far it has been great tell us about your first first failure why was it so catastrophic what did you learn from it uh it was so catastrophic because I feel like I was just a kid uh super naive um we the company was an advertisement company so we would take like TV screens and put them in elevators and then sell those ads to companies and so we would go to like these massive buildings that had like four 400 apartments or five thousand uh tenants and we would sell the ads to those buildings and that we would find like locals who wanted to add uh advertisements there but as you can think about it like the operations of something like that um is significant like you you have to change the ads every two weeks you have to maintain these screens you have to clean them you have to sell you have to sell also to the tenants so I I think there is a couple of little businesses like that in Latin America it's definitely a model that works but we were just not uh in a position to uh build the operations engine required for to manage something like that uh so it was quite painful um I remember I had to do a bunch of hacks just to upload the data into SD 5 SD cards and uh I never thought about scale until I I faced that problem because you say like oh you can do that like in a couple of minutes right but now you multiply that by a thousand a thousand uh SD cards so how are you gonna upload ads different ads to different cards um it's just a really big problem and I had to do all of the technology there uh at the end we started moving into like 3G and other types of Technologies to to upload data but it was just too late I got super burned out we were not earning a single dime um so I decided to to give up basically interesting yeah I've had various of these moments myself but I've been always like super boots up and obsessed with not having a job somehow it's mostly an obsession rather than a rational thing because you occupied various jobs at various big companies and not for a long time I guess you've always been like side hustling um but yeah Shutterstock customer data dog which is hyper popular nowadays meta I want you to tell us what you learned from these companies what insights you had and am I right in saying that you took like the the big paychecks you earned there you save it up to kind of cover up for potential rainy days for your next uh big Venture your next big entrepreneur uh Venture there yeah I uh you're absolutely right I I I'm a person that is really good uh living below I mean um so I've been saving money uh since I started basically um I I basically only spend money on travel that's my one one thing and when I do it I really enjoy it um like we were talking before we started I've been to Mexico multiple times uh I absolutely love that country I love traveling meeting new new cultures so once I accepted that I was gonna spend money in the this one thing and and the my rest the rest of my life is okay as it is uh as I keep earning more and more I was I was totally okay saving um and that definitely helped me make the jump to starting metal because I actually quit before um joining or having joining YC or having any money at all uh invested in the company um and now going back to your first question uh I learned a lot about teams and how they change depending on the scale that they work in so I started in a company like customer when we were only about 20 employees the engineering team was really really small and uh I also noted noticed that the velocity changes a lot in types of problems changes the company grows and even when I moved to uh the the problems were so different and the philosophies were different and so like learning to adapt and figure out where your best uh where your best effort is is the best skill that I I probably got from working in all of those companies right just for the audience to understand working at meta the salary range in that position is it fair an accurate uh estimate to say that it's at least 300K a year um I think there's a website levels either gives you like the ranges but like the ranges are crazy like you can you can have like an entry developer earning 200 um and someone with 15 years of experience earning a million right um and and there's a lot of like variable the variability to the income uh so if you hit like some marks you were getting like multipliers uh which affect your bonus and then you also get Equity refreshes it's kind of like a Snowball Effect uh so you you feel like you have to say and you have to say and it keeps you it's it's worth Golden Gloves uh come from right uh it's definitely true right and I mean you live in SF and New York City two of the most expensive cities in the world but you often reflect like uh with your friends or family back home what does 300K a year mean for them when they would kind of gain that in a lifetime do you often think about that yes I do um and it's it's something that I've always discussed with my wife actually um we've been discussing just moving back to Colombia uh where I grew up I I think that there will be a moment where I will make that happen but um at the same time I do feel like right now being here and in San Francisco is the place the best place I could be to improve my odds of success um and and yeah the the difference in pay is so astronomical it's really hard to justify um I do think that the the best people that I worked with are uh people that I've worked with in San Francisco and New York City but that that's not to say that there's really great engineers in Latin America I know there's companies like rappy and Colombia that are generating really really uh talented Engineers a lot of them that I uh I have met over the years um including some that were in YC winter 23. um and so I I think it's just part of the opportunity cost as well like uh living paying rent in in a place like uh New York City could be for a one bedroom upwards of 4 000 a month so that taking into consideration with the that taking into consideration plus all the um the taxes that that you have um it still is a really significant difference um but at the end of the day is I think it's about the opportunity cost and and getting the best people that you can um yeah right there's also Austin that a lot of people are talking about Miami which is a nice lifestyle as well and lowish taxes um indeed I I do think well rep is one of my favorite startup I think that YC has started investing in latam startup quite uh heavily and other companies as well which is uh so cool Colombia is one of the best plays on Airbnb to get a high quality airbnbs for cheap alongside with Chile interestingly enough I do think that the real estate market is somewhat of a bubble but it's a bubble that we all believe right it's like oh I invested in this and I don't want to lose out so may I have this fantasy of like building houses for 500 bucks you know like it that's one of the moonshots I'd like to do um one day uh but let's talk about uh remote work do you think meta would let you work from a distance or do they only pay these salaries for people that actually stay physically and go to the HQ and tell us about the advantages vs disadvantages of remote vs physical work yeah it's a big topic so I think that uh during the pandemic they had like a pretty open uh philosophy on remote work so you could become like a remote worker um in that that would basically make you give you the freedom to work anywhere in the U.S um and if you wanted to work from outside of the US you have like some certain number of days a year where you could spend that time uh anywhere in the world uh obviously with some limitations like I think you you couldn't work from Russia and in a couple of places like that um but and now they're asking people to go back to the office I I believe in for example my company is has been like in the office every day so I think that depending on the stage of the company and the type of company that you want to run uh remote obviously makes a lot of sense I I see uh the company that I work in that I'm working on becoming eventually remote friendly but right now I value the time that I spend with my co-founders every day in a room um I value that more than anything else and and just to give you an example when I started YC we got an Airbnb with three bedrooms uh our our wives and significant others moved in with us and we spent there three months every single day working uh we would wake up and work until we went to bed uh we had we had Saturdays off at least but Sunday still Friday nights it was just pure work and obviously that's not sustainable it's not maintainable but I thanks to that uh exercise I was able to Crunch a lot of productivity get to really know my co-founders and and figure out a lot of problems that we had at the beginning with the idea we pivoted a lot um which is also something that I figured out um Midway through YC a lot of companies and white combinator pivot Midway or even at the end so we had a goal of not pivoting uh by the end of the of the batch and and I believe we succeeded with that um so that that's why remote works where you start scaling your company where a lot of the work can happen asynchronously but at the beginning I truly believe that being in the same room does matter yeah I also think so I'm gonna try the physical work like me it's weird I've been an entrepreneur for the past decade and I've never done physical work most people that I meet for the first time physically it's because we do like employee events or like meetings whatever you want to call them um but yet like quite impressive that information just provided there and yeah I can see the non-scalability of it with your wife you know like uh it's like a bit privacy and it's like man they're so obsessed you know but they must also understand in some way um I'm curious to talk about the non-pivot thing which is really uncommon especially in YC why why did you fix the gold to not pivot and did you tell basically your your advisors that and they did they agree um yeah yeah they actually did I I think that a lot of people see what I see as this Pinnacle which I I think in many ways it is right it's it's like it's the best accelerator in the world um only you one one to two percent of people get accepted um and so you see this place as like where really much care companies arrive and achieve the the goal of scaling bleats Blitz scaling uh and and do millions and uh what we noticed is that it's it's a place where companies go to ask questions and to learn um and to really figure out where product Market fit is um and a dirty secret is that nobody has product Market fit you know uh a dirty secret is that nobody knows what they're doing you know and so it's it's about accepting that you don't have all the answers and and uh not even your partners have all the answers you know so you you just have to deal with each data as you go and when we arrived it was like an ego death in the sense that we we were saying like oh we're gonna crush this we're gonna be the top one percent we're gonna kill ourselves and and nobody will will beat us you know we were thinking like that and then two weeks two weeks in we're like oh what are we doing and we started having like these really hard discussions like we honestly don't know what we're doing we have this idea we have this know-how we know how to build anything but you just feel um it's just a hard problem like figuring out what to build and what is the best thing to sell it's so complicated that um it's just uh it destroyed it destroyed our ego in the best way possible and so uh after those couple of weeks that we were like figuring out what to build we we found like a really good idea and you stuck to it we we stopped thinking about just beating everyone and and we just focused on ourselves and and that mind shift was really what changed everything for us uh we just started opening more to our partners we started uh admitting when we really were not sure about something and and avoid those like insecurities or just confronted those insecurities and um and yeah so but but that uh at that point we just decided like let's just focus on building something uh people want let's talk to people every day does not think about like Blitz scaling and and having millions in sales and just focusing having something that gives us enough security about the future and and not pivot not change idea just if we achieve having that commitment to an idea then we are successful and and and so far we we've that's been great like um we've we've changed our minds a lot about a lot of things and now we're just focused on building and building and building and learning as much as we can from our users and customers right who had the idea because the idea is so good and you know well a lot of people nowadays they're like oh ideas aren't worth but the thing is once you get a good idea first it takes a lot of time and mental work to get that idea product Market fit and you know it's kind of quantum physics but once you get a good idea and you can try to scale that idea well it helps a lot to have had that idea in the first place not sure if that make any sense but who had that idea can you explain uh to it to us like what uh is metal exactly and and why the the idea can actually scale yeah absolutely so when we arrived to um to YC we had this idea that was to uh gather a lot of customer feedback from multiple channels emails tickets GitHub issues gear up tickets like any any surface uh area that was possible and then abstract away some insights about what customers were saying to kind of feed the the product manager with ideas of what to build next so that that was a that was our business right and the technology that we were going to use was embedded and uh I'm not I'm not sure if you're familiar with embedding but it's it's a way of representing the meaning of races and text images uh in numbers so you can find similarity how how things are similar or dissimilar to each other um and like a couple like a weekend we we we thought like wait this is really hard to sell um are selling to the product is really hard to product managers they have so many tools already that is a big challenge to add another tool to that tool build is not a harem harem fire problem either and so it we we decided just to start exploring all their ideas and we got into for example how to build a better slack so that was like our next yeah that it took like a couple weeks to disapprove we got into like Integrations um and then I had a conversation with someone from open Ai and they just brought up embeddings and I was like yeah we're pretty familiar with embeddings we were going to do something with it um with them and and all of a sudden like it clicked and I was like oh why don't we just build infrastructure or uh sorry I guess yeah no problem I'm sorry I'm sorry I thought you were you left my bathroom sorry about that one second no problem yeah I went over time since my band no no issues but uh yeah I mean finish your last uh your last one because uh yeah I understand your time is quite precious as well so to finish the the open AI inside is quite uh interesting I also had various conversations with open AI folks my I mean it's the most impressive company in the world and talking to people certainly helps it's like somewhat the basis behind this podcast right um so my understanding is that you kind of iterated on the idea with your past experience plus what's happening in the current market and you had this idea you spared mentally with your co-founders and you're like okay let's do this and then you went into yc2 to kind of test the hypothesis yeah exactly and um yeah when we so when we talked to uh someone from open AI um I started thinking like there's all these companies that are um starting to build things around uh uh generated AI one of companies that are moving into this I don't want to be one of those companies immediately but I kind of want to work with infrastructure we really understand infrastructure and developer tooling so our bet was immediately focused on helping those companies get the product faster uh because there's a lot of setup for a lot of those uh problems like you have to set up your own Vector database you have to set up your own API to consume it your pipeline to index all that data generate the embeddings so there's a lot of pieces that you can definitely build yourself but if you can save a lot of engineering work then people will will pay for it and that's that's currently our bet right so kind of sell shovels and picks and jeans and helmets to miners rather than build their own mind yourself which is pretty logic and pretty yce um let's keep it to this today Sergio really grateful for your time I know that yeah you're you're like you need to put your head down in code today as you usually do thanks so much for the insights working people find out more about you and Metal yeah thank you so much uh you can follow me on Twitter my handle is Sergio products 666.