Jake Aaron Villarreal: Welcome to our podcast, From the Ground Up, where we interview startup founders exploring their journeys, their challenges, successes, and lessons learned. We hope you'd be inspired in discovering what it takes to build a thriving startup. I'm your host, Jake Aaron Villarreal, and here with us today we have Zach Rattner, the co-founder and CTO of Yembo AI. It's raised $12.9 million in funding and it's grown from 9 to 70 employees. He also released a book called Grow Up Fast: Lessons from an AI Startup, a fascinating read, hope you enjoy the episode. Zach, welcome to the show.
Zachary Rattner: Hi Jake. Thanks for having me. Good afternoon.
Jake Aaron Villarreal: Yeah, great. Well thanks for joining. Um Zach Rattner is a CTO and co-founder of Yembo. He's got 15 years of experience in software engineering as well as a BS in Computer Engineering from Virginia Tech. Yembo is the global leader in AI-driven virtual surveys, and Zach holds 19 US patents, is in the top 2% of Stack Overflow contributors, and recently authored Grow Up Fast: Lessons from an AI Startup. So we'll touch on not just the company you're building, but also the book you wrote. Excited to kind of walk through that a little bit with you. Before we dive in though, uh give us a little background of you and how you really got into the startup business game.
Zachary Rattner: Sure. So I'm an engineer by training. Um I've always been kind of a nerdy kid. Then I discovered math, got good at that. Um went to school, majored in engineering, and then worked as an engineer. So I've always been I would say a technical sort of person. And I'd always been kind of interested in startups but I didn't grow up in the Bay Area. Uh my family moved around a lot. So I always at some point wanted to do one, but I also didn't want to just go straight there and have it kind of not work out. So I ended up um interning at a startup when I was working my way through college. Then I wanted to go get a big job at a big tech company to kind, kind of learn how engineering works at scale. And then um after about 5 years at telecom giant Qualcomm, I ended up deciding it was time and then switched over and um have been at Yembo ever since.
Jake Aaron Villarreal: That's great. Now is Yembo your first startup or have you had a few before this?
Zachary Rattner: It's a bit of a trick question in that I have used the Yembo name for every side project I've ever had since I was 15 years old. Um so yes Yembo is the first, but it has had earlier iterations. I would say this is the first serious one. I mean when I was uh 15 years old um designing people's websites after I'd come back from my job at the grocery store, that was technically called Yembo. But I didn't have a company incorporated, I didn't have like full-time ambitions even, so I would say this is the first time we've done it for real.
Jake Aaron Villarreal: That's great. I love the name by the way, so I could see where you would just transition that to the next company and you know, it kind of sticks, you remember it. Um when you uh started building the company we're going to first dive in a little bit about your company and then we're going to hit the, the book that you wrote, sure, because it's really, it's easy to read but it's also, you can learn a lot. And especially in today's world around AI and what's happening, where it's being applied, um some of the lessons that you've learned and as you continue to scale your company, I think it'll be really insightful for others to listen to that. The company you started today, what inspired you to build Yembo? What's the story behind that and what problem are you solving in the market?
Zachary Rattner: Sure. So the, I'll quickly go over the product and I'll tell you how we got here. So in a nutshell, if you are looking to purchase a home service, be it property insurance or moving, traditionally that's really difficult to get a quote because somebody's got to come to your house and look around and inspect what it is [that] needs to be done. And we use computer vision technology that accelerates that. So rather than needing to have someone come to your house with a clipboard and if you're moving, identify what items are there, how many boxes will it take to move, we can take 10, 15 second videos of each room in your house and our AI will identify what's there. And we sell that um as a service to home service companies, moving companies, and more recently we've added property insurance as well. And it provides a much easier, faster, and more accurate result because you're able to actually have like the pictures and the visuals of what was there. So that's the product in a nutshell.
Um, you asked about how it came to be. It's kind of a weird use case for computer vision, um and that's a bit by design. So I was working at Qualcomm on this new te- technology team. So Qualcomm traditionally has had its core strength in the modem chip, so the part of your phone that talks to the base station of the cell tower. Most major handset manufacturers like Apple and Samsung, they buy parts from Qualcomm. And um as mobile computing got better they started looking at newer areas of um not just cellular technology, but I mean they also have a graphics chip and they do a lot of high performance computing. So I was working on a team that was looking at emerging areas and um computer vision specifically was the emerging area at the time that we were looking at. And I would go out to all these trade shows and conferences and learn like "what is the state-of-the-art in computer vision?" And it was like mindboggling what was becoming possible. This was back 2015 or so. Um self-driving cars uh were uh kind of hitting the, the mainstream. People had been working on them before but it was like making enough progress where people were starting to really understand this is like something that you can use in the near future. A lot of applications around drones. And there were just so many things that were fundamentally changing because the technology had gotten to the point where computers were better than humans at identifying objects in images. So there was a popular academic benchmark and some threshold was crossed where now this very human task of "look at something, tell me what's there," uh computers got better than people.
My wife was working at a moving company at the time. And um we, she was working remotely and I would come home from work and our living room, the coffee table would become the back office of a moving company. So I got to see what um it takes to really make the logistics work behind the scenes. Um and she worked in an international department as well, so that's probably the most complicated division. I mean they all have complications, but you have to deal with like customs, foreign languages, different time zones, a lot of things that if you're just doing like a two, three mile move down the road you don't have to deal with. And that was really the genesis, was realizing that here's um this industry and it's just very reliant on pen and paper, email, written down notes. Um and it's not for lack of trying. I mean a lot of these companies uh they have like very high profile clients, and moving is important, everybody does it. About 40 million Americans do each year. And so it's not for lack of trying, it's just it's very difficult to do a good job. So I was um living in these two worlds kind of trying to reconcile how can we be dealing with self-driving cars on one hand, but on the other hand if I want to know [if I need a] 12 foot or 16 foot truck that takes like 3 hours of manual labor to, to give me an answer. And uh one thing kind of led to another and uh and then Yembo was born.
Jake Aaron Villarreal: And she didn't just work for a moving company. I mean if, I think we had a quick conversation and just correct me if I'm wrong here, did she help, did their company help move the president into the White House? Is that the same company?
Zachary Rattner: Yeah. So it was a, it was a Washington DC based company. I forget, I forget which president. I think it was, it might have been Obama, um but they had um, they had contracts with a lot of um high ranking government departments. So pretty much every Secretary of State they probably moved. So it's like kind of cream of the crop, if there is a cream of the crop moving company I would say that that's it. And um it's just very difficult to do these kinds of jobs and to do them well. So that was, that was the one thing that um kind of really convinced me. Because it wasn't like it was like a, two people in a pickup truck and they're having a hard time. It was, we learned those people are having a hard time, and everybody else all the way up to the biggest and the best are also having a hard time on the same space.
Jake Aaron Villarreal: So if you're a user of this, so you're moving or you're going in to survey the house, you have a lot of art, maybe you have big furniture pieces, you're moving cross country, rather than moving everything over and then finding out that your couch doesn't fit in the family room and your pieces of art you expected to have in the family room aren't going to fit, you can literally walk through with a mobile app or a phone and just kind of take video and it will tell you what will [fit], what won't fit. Is that kind of the, the problem it solves?
Zachary Rattner: Yeah. In a, in a nutshell. So the, the biggest difference from what you said is we work entirely in the browser. Um we found that people are shopping around, they're not really committed. So I wish we could have done a native app, it would have been a whole lot easier, but to, to make that all work in the browser was much more complicated. But we come up with an item list. So you take a video and the AI does a few things. It summarizes that down to a few images, so rather than having to watch a 30-second video you'll see like maybe five or six images. And then it'll identify the items that are in there. So we show you just enough images to kind of cover here's what's in, in here. And then we identify each item and we give you a size, volume in cubic feet and a weight in pounds estimate which has just kind of been honed over the millions of objects that we've detected and learned from over the years. So that end result, you're not just seeing "Oh, it's going to be a three hour move," you'll see "it's an 8,000 pound move, 1,200 in the living room, here are the different items, here are the packing materials." So everyone can just be more prepared on the day of the move: the crew, the customer. Um then when you're unpacking you kind of know exactly what you're getting yourself into because you can see the entire list.
And we started thinking it would be a tool to kind of help maybe resolve disputes, and what we found is a lot of times it prevents that dispute from even happening. Um because I think there are bad actors on both sides. There are bad consumers, there are unscrupulous movers, and when everybody knows that there was a visual that's date, t-, date stamped, timestamped "this is what we agreed to" um a lot of the claims around like "my 11th TV went missing and I need a damage claim payout" just kind of like disappear when, when people know that there's honesty on both ends. So we uh, we bring transparency and we bring a better customer experience that takes less time and less energy to go and do.
Jake Aaron Villarreal: That's great. Is it a product that you're selling to big service providers like moving companies, or is it to a user that's looking to make a move and you just want to kind of validate like, "are my pieces going to actually fit in that new location?" Who are you selling to?
Zachary Rattner: Sure. So we uh iterated on kind of every different version in our earlier days, every different idea. And what we settled on, like a lesson I think we learned from this just in general, is you want to um charge the entity that gets economic value from working with you. Um and the moving company in our scenario has a real hard economic cost for not having this. If they have to drive, there's wear and tear on the vehicle, the hourly rate of the person, like car insurance, fuel, all these kinds of things. So if you don't have Yembo, we're able to do a study and see it costs anywhere between like $250 to $300 or so to go perform one of these estimates. So going to a customer who's not committed to do anything yet, they're just shopping around, saying "hey can you pay me some money and I'll, I'll tell you the estimate," it's just a much harder um harder proposition. I think there's still value, but you need to convince someone who's maybe not themselves convinced that they have a problem, whereas the mover they know they have a problem. They have a certain way of doing it. And that time to value is uh much, much shorter than if you try to just like buy Google ads and hope for the best.
Jake Aaron Villarreal: Makes sense, that's great. So you started the company and COVID comes. And COVID's a time where for comp-, some companies that really helped them, and others it didn't. How did COVID impact your company, and when you started then, or prior to then and where you're at today, what was the company size then and what's it like today?
Zachary Rattner: Sure. So I think when COVID hit we were low double digits, maybe 10, 11 people. Um we are about uh 70 worldwide now. About 40 in the US, the other 30 international. Um so I would say if you stop right there and zoom out, COVID did uh helped our business, but I would say it wasn't a straight line and it certainly wasn't um like all done for us. I think there was a lot of momentum that it gave, but initially what happened was it accelerated a lot of things. So if people were on the fence, they would either freeze in their tracks and say "now is not a good time to change, there's too much change going on." So in the like early days we actually lost a lot of prospects in our pipeline of people that were evaluating it, it was going well, and then they just kind of froze and said "Okay, this is, this is weird. I don't know what's going to pan out yet. I don't really have the appetite to retool my sales process in addition to dealing with shelter in place and all that kind of stuff." On the flip side though, um it's now much more uh costly in a sense to go into somebody's home right? Because there's, there's, there are lockdowns and people aren't always open to it even if you are considered an essential worker like a lot of movers were. Um so in that case it did accelerate. We had some clients that were trying us out, we had I think a, a 12-month pilot with a large, larger company that got condensed into like 3 weeks um because shelter in place happened partway through. So I'd say all in all it um it helped accelerate, but it really took um understanding the skittishness that was going on in the market. And then we built a couple newer features that made it easier for customers to use it. And I would say like all that said and done, in hindsight it, it helped, but um we couldn't have just like taken a two-year vacation and had it all work out in the end. There are a lot of, lot of things we have to do along the way to retool on our end as well.
Jake Aaron Villarreal: Exactly. You talked about your product, and I know customer feedback is important and being able to make your product better. You know I recently read the book, the recent uh biography of Elon Musk. And he talks about over-automating the Tesla uh manufacturing facility. And you know, he uses the algorithm, which if you've read the book kind of walks through how he takes something, rips it apart, simplifies it, and then makes it you know lower cost and more efficient.
Zachary Rattner: Yeah, add back 10%, then you've not cut away enough. Exactly.
Jake Aaron Villarreal: Yeah, so you're really looking at asking hard questions and questioning everything, ripping it down and building it back up in your product and the company that you're, companies you're serving. Did you go through any of that? Were you... tried to automate everything or make it simpler, but the customer maybe thought differently? Any, any scenarios like that that happened as you were looking at optimizing your product fit to market?
Zachary Rattner: Sure, yeah. And I think the biggest thing to that is that uh I would say our product kind of pioneered, and we had to get over the home... like pioneering I feel like sounds really good in theory and people like to talk to innovators. It actually makes your life a lot harder because nobody knows how your thing really works because it's not really like anything else. Um so there are some days I complain I wish I was like selling Outlook to a bunch of Gmail users and you just kind of go down the features "tick, tick, tick" and [on] your way you go. But with our product it's a workflow change right? You can't just turn a switch on and have this, you have to have people that um used to drive and ring doorbells and they need to come and send Yembo links instead. And then when the results come back you have the ability to quote in such a little amount of time, but you need to like quote it. If you need to look at the results, you need to get back to the customer. So if you don't have a good process in place around Yembo, you won't see good results.
So I think the biggest thing we learned in the early days was, I'm an engineer by training, my co-founder is also an engineer. And we just thought, um wow if we made this possible everything else is like a done deal. And what we realized was there are a lot of adoption things that you need to overcome where you build this and it is better, like customers like it, the moving company likes it, you're able to do more business. Um so we have certain clients that have been able to like double, triple their sales volume because they're not spending so much time sitting in traffic or asking people how many boxes will it take to pack these books. But when you say, "hey salesperson who's been doing this for 30, 40 years, I'm going to use AI and help you," I don't think movies help, but like they think Terminator, they think HAL from 2001: A Space Odyssey. And um they, there's sometimes a hesitation to adopt because you're worried that like the robot is going to make the machine ir-, or make the human irrelevant. And I feel like that's a very um kind of like Hollywood based, maybe rudimentary view. And a lot of the applications that we see, what we provide and also like all these tools that are out there... AI is a tool that gives a person superpowers, but you can't really take the person out of it.
So there are certain things that computers do really well. I don't like adding up 10,000 numbers. I've got Excel to do that for me. But that doesn't mean that like you don't need a person because you have Excel. And at the end of the day AI software, it's still software, it's still math, it's still science, and um at the end of the day it's a tool. And there are certain things like building rapport with the customers, um being friendly, being responsive, trying to understand really what's important to that client and then providing that service. The AI is not going to be able to do that, but it's great at counting um how many boxes, it's great at identifying things and taking the maybe more tedious parts of it out. But people think about extreme cases. Um so what the tool actually provides is: imagine the kind of conversation you can have if you're not counting boxes. You're just building rapport and u- getting to know each other, understanding like what's needed. And the AI is there kind of figuring everything else out for you. Um so that's kind of I feel like the big misconception, was that we tried to um just build this like amazing AI was going to do everything. Then we realized there are actually certain areas that AI is just not the right tool to do.
And that didn't just affect the product, it also affected the positioning around it, and um making sure that clients who are using it are like getting good value out of it. Um and as engineers we care a lot about like, "oh this is a cool technology, it's something that's never been done before." And there's almost a sense that something is inherently better if it meets those boxes, but in business a lot of folks don't care about that. They have a job they want to get done, and they want the most efficient and effective way to get it done. So just kind of um being open to feedback, not assuming that you've nailed it just because you've finished in your mind. And actions do speak louder than words, it's a cliché I would say for a reason. So once you've launched, don't just listen to like people saying "how do you feel?" also look at actual outcomes that are coming from it. How are you closing business? How long does it take you to get back to a customer? We had a few hard metrics we actually looked at and monitored and made sure that like, if this thing is really going to be as valuable as we think it is, I should see these outcomes come from it. And that was kind of the lens that we filtered our feedback and iterated the product based off of.
Jake Aaron Villarreal: That's really good. It's sometimes scary to listen to feedback because you don't want to make changes if you think you're right. But ultimately it's the customer right, and you have to adapt to what they say. You know, even in our industry, you know AI is impacting a lot of different sectors, but you know we recruit for companies, we help them find the right people and scale, and we've been using AI for a long time. It just hasn't been published at the, you know, in, in a way that it is today in different areas. Um but you know, we can go through and do our business 70% better, faster and we can push that cost um out of our equation and [create] better offers for our customers. But ultimately it's a people business. You know, at the end of the day you can get stuff done and automated, but when it comes to making a life-changing decision, you know people want to collaborate sometimes and think through "is this the right place?" or "X might not know of an opportunity." So I think AI uh to your point yeah, will certainly help in a lot of areas and make people better at some of the mundane things they have to do, but ultimately you know there's humans behind the business. So for sure, I, I, you know, people, a lot of people are afraid. But I think, I think you have to embrace it.
And so I'm really excited to kind of talk a little bit about your book here. Because you know to go from building a company and being a startup founder to being an author takes time. And it also is you know something that um could be very valuable to a lot of others if it's specifically talking about things that startup founders go through. So your book um is uh interesting to me and I have some questions about it. I, I want to kind of ask a few and, and we can go, go through that. But um you know the title of it again is Grow, Grow, Grow Up Fast: Lessons from an AI Startup. So in terms of the book itself, before we even talk about the book, what gave you the idea to want to write a book?
Zachary Rattner: Sure. So a few things happened at once. Um there was a lot [of] change, a lot of change happening in the AI space. I mean even if you look at uh the world we live in today versus the world pre-ChatGPT, that was less than one year ago and it feels like everything's kind of been infiltrated. So I think the, the pace of acceleration in the space had kind of picked up. Um and I also don't have a fantastic memory. Um I love reading books from uh former founders kind of later on in their career. I think like Peter Thiel's Zero to One or Richard Branson's The Virgin Way. They're both really entertaining reads, you learn a lot from it. But I feel like if you're looking 20, 30 years back I might not remember all the lessons that I've learned um because it's been too long. And then also a lot of the stuff that for instance like Richard Branson talks about in The Virgin Way you can't really do anymore. Um he talks about like uh breaking into airports after hours and painting Virgin logos over British Airways. Like that's a great way to like uh spend the rest of your life behind bars.
So I wanted to kind of encapsulate the things that I had um, that I had learned. And I found in my day-to-day work and with like other founder friends that I'm mentoring, I found that I was repeating certain lessons over and over again. And a lot of the engineers at our company now, they weren't here when it was two people in a garage trying to get like a scrappy fledgling thing off the ground. So I wanted to kind of consolidate everything that I had learned and share it with a wider audience at a point in time where I felt like there was like a, a gap and a need in the messaging. Um and it ended up taking probably 10 times longer than I anticipated. Um I think we can maybe talk about that a bit. But uh there are just so many details that it takes to do a good job, and it's sort of uh, I see some parallels between writing a book and, and writing software. Um the first draft I was done with in about two months. And I had an outline, I got, I worked with an editor, he um kind of worked with me on the outline, he held me accountable to word quotas certain amounts each day. And um I got to the end and said, "Cool, I'm done."
So I take the manuscript, and just like with software you beta test it at that point. So I went and I found beta readers in our demographic. And I was so proud because I got the whole thing finished and um my artist had finished the cover art so it looked great. And um one of the questions in the feedback form was, "If you had to rate what you just read on Amazon today, like assume no other changes, what would you rate it?" And I came back with a whopping 2.5 out of 5. So that was a bit humbling. And what we learned was people had very different expectations going in. There were people that were engineers by training that were saying "it's not technical enough, there was no math, I didn't have to like bring out my calculator to do anything." We had um other folks mention uh that the challenges, we kind of glossed over it, it was too easy. We talked about the outcome, but we didn't talk about the like alternatives or other things that had come up. And it was really helpful I think. Like listening to clients, listening to customer feedback, iterating on it. So the version that you have in your hands now, and the one on my desk, this one got a 4.9. I figured uh you can't uh you can't hit five, that's fine, I'll, I'll accept that. But um that's kind of how we, how we launched it. I've never done a book before, so I just kind of wrote what I felt the audience would want. But then those assumptions get tested and you get feedback, and that's why it ended up taking kind of a long time. Where it wasn't just like u- I'm not, I'm not Dan Brown so I can't just like hand it to them to do it and it's done. You kind of have to like test it out, see how it goes, get feedback, rinse, repeat.
Jake Aaron Villarreal: Got it. Well, I'm going to ask a couple questions on the book and you can, you know be great to just get some insights about without having to read it, the listeners can kind of hear from you directly. So one thing that I thought was interesting was you talk about Radiohead and how Radiohead had an impact on you and the music of building what you create when you code. Different elements of music and tying into how you operate. Walk us through that, because I know there's a lot of fans of Radio[head]... I'm one of them, I'm a fan of Radiohead. It has a big impact on you. Walk, walk us through that.
Zachary Rattner: Sure, well as, as a fellow fan I think you'd, you'd uh you'd appreciate this. But it's not the kind of music you're going to find at a club. Or if it is, let me know, cuz I'd like to go to that club. But it's um, it's almost like a paranoid music. You're thinking about things that could go wrong. There's a lot of talk about um the uh, about um kind of the dark underside of people. And um even if you listen to like the song structures, there's uh, it's not just like happy-go-lucky four-four time, it's... uh I read, I read a review of um, of one of their albums that said that Thom Yorke is the king of asymmetry. But like it works. But it's uh, if you look at, listen, break down the song "Paranoid Android" and look at how many um like tempo changes there are and how many key changes there are. It's um, it's complicated and it's like weird, and it kind of makes you think. And that's been um, a few of their albums have been my favorites for probably 20 plus years now, and I still listen to them sometimes and like get something new out of it.
So it's um, I feel like it's a good proxy for what you're doing when you're putting a startup together. Is that to an outsider, startup life looks like the hockey stick up and to the right. Like you have your seed round and your Series A, and everything gets bigger, bigger, bigger. But real life, it's got layers to it. It's got complexity. It's got um, you make motion but you're not sure if it's progress. And I just feel like the um, the tonal qualities of Radiohead and the lyrical qualities, it just, it's a, it's a great um kind of uh proxy in a different industry, in a different venue, but like similar, similar concepts.
Jake Aaron Villarreal: You know I traveled around the world uh on my own with a backpack and a camera. And Radiohead was the theme behind my travel. And I remember sitting in airports looking out waiting for my plane to come, listening to Radiohead. And just had a really good memory of that in the background of you know, it was almost like a movie traveling, so really cool. Um your friend at the dinner table test. What is that?
Zachary Rattner: Sure, this is a test that we've used a lot at Yembo and it is a great gut check um that we found. Because in startup life, most of your team is going to be doing work that they've never done before in their career. Which is great, and that's usually how you recruit the right talent, and it's um how you can make progress and, and do awesome things. The downside to that though is it can be kind of hard to know if your expectations of somebody else are reasonable or not, because I've never done it, you've never done it before, we're trying to figure this thing out. So the "friend at a dinner table test" is a pretty simple thing that you can do. And uh it's a little bit of a role playing exercise. So let's uh let's set the stage and then we can, we can chat about it a bit.
So let's say um you are a new engineer on the team and I have this first project that I want to give to you. And I'm pretty sure it can be done. Um I know how I would do it. But um there are some, I don't know, external factors or some reason that I have a like a timing deadline in mind, and I'm not sure like if it can be, if that deadline can be met or not. So imagine I give it to you. And for the sake of argument, assume that that conversation doesn't go over well. So you're super annoyed with me, this company is unreasonable. They uh, the worst case possible scenario happens in that conversation. Then the workday ends and assume that person is talking to their friend at the dinner table. And because you just um annoyed them, they're going to be complaining about you naturally. And the exercise goes like this. So put yourself in that situation, and then you can grab a friend and roleplay either side of that. So one person plays the role of the annoyed employee, the other person plays the friend, and see if the friend will try to corroborate or if the friend will take the side of the company.
And um I have found that that's been super helpful in terms of um framing things, in terms of like explaining how something is going to go. And um it actually convinced me to promote one of my engineers to a management position. Um this engineer was doing great work, had never managed before. There was some debate internally with the executive team um if the person was ready or not. So we, I did the friend at the dinner table test, and um because I thought we had a good thing going. This person was a great engineer. If I like push on the gas too hard and they hate it, I don't want to like um break something that's going well. And um and the line that the friend said was uh... I called one of our investors and I had them be the friend and we're going through back and forth with this. And the line was uh, "Come on man. So here's what you're telling me. That your boss believes in you, wants to give you additional responsibilities, and is willing to help you, willing to give you resources that you need to do that. And you're, you're concerned about what it might possibly give up that you don't even want to explore that and see how it goes. Just try it out! And worst case, if you hate it, then if you have a good relationship like you said, tell them you hate it and see if you can switch back later. Like why would you, why would you not take this opportunity?" And um that kind of uh put the thought in my head that uh maybe we should be doing this.
And um I think it's been a helpful test because [in] reality that situation played out pretty darn close to that exercise. Um and it's not perfect right? You could game it. I could take like a super adversarial attitude and like do ridiculous stuff. But I feel like the point is that you're forcing yourself to think with empathy from the recipient's position. And then once you understand not just like from the boss's perspective what you want, but from the other person's perspective how it's being received, and what the alternatives from their perspective are. I found it just to be pretty helpful in terms of um making it very like concrete. Um you can see as a, as a software engineer I like to iterate and test. And if we can have like a fake conversation that doesn't really matter, with low stakes, off on the side and then come back and do the real one later, then I found that's pretty helpful.
Jake Aaron Villarreal: Oh that's great. Really cool. Um The Time Warp. What's The Time Warp?
Zachary Rattner: The Time Warp is my attempt to help people understand [that] time zones get a bad rep. I mean, if you hear people complain at work, "oh the team's in the other country, so someone has to stay up really late," I feel like if you do time zones well you can make 24-hour progress and nobody gets burned out because nobody's working 24-hour days. It's just the company is getting work done 24 hours a day. So what we did at Yembo was, we took the workflow that it takes to design a new feature, build it, test it out, launch it, and we broke it down into three discrete chunks. There's like the initial system design, architecture, scoping. And there's a lot of coding and implementing. Then there's integration testing and making sure that it works on all different browsers on different iPhones, Android phones, iPads, all that. So we have a team in the United States, we have a team in Ukraine, and we have a team in India. And if you look at where it is 9:00 a.m. to 5:00 p.m. on a world map in those three places, there's very little overlap. And that's kind of by design.
So now, you come to work in the morning, you start working on your features, you scope everything out in the US. Then we have an internal project management system where the team in Ukraine kind of picks it up and does implementation work. They say they're done, the test team picks it up, tests it. You come back into work in the morning and progress has been made. You can review what [has] come in, you see the test report, you say, "Cool, it's ready to go." So it takes work. You have to know your business and your process and your workflows. And you have to kind of engineer the handovers so that people aren't needing to have like five-hour Zoom calls just to hand stuff over. But if you pull that off well, you can make 24-hour progress. That's kind of the, the, the pot of gold at the end of that rainbow. Is if you do it well, you can, you can keep things going. So we have a few examples at Yembo where that's kind of saved us. But we've um, we put it in place around 2018, 2017, pretty early on. And it's still, still humming along today.
Jake Aaron Villarreal: That's great, a couple little nuggets in there. Lessons, I love it. Um you know, we're in a world right now in an economy that's, there's a lot of uncertainty. In terms of what the news says. And you can buy into it or not. But ultimately, you know, there's a lot of anxiety for people and companies and leaders, and forecasting what may or may not be coming. Managing uncertainty, talk about that. How do you do it?
Zachary Rattner: Yeah. I think we talked about this line earlier, about "everyone's doing something for the first time um at a startup," and managing uncertainty is, is going to be part of it. And I feel like if you look at the students who get 4.0 GPAs and As, they are usually very good at like being prepared, um having done these "I did my homework, I did well on the tests, and now the final exam's going to come around, I did a good job." All of that kind of gets thrown out the window um when it comes to a startup, because you can't study. Um I mean you can do certain things to be prepared, but at the end of the day um there was no like "global pl- pandemic preparation course" that came up. When COVID hit, we just had to kind of figure it out.
Um so I think with the, with the un-, uncertainty that's in the market, and I feel like there's always uncertainty, it's just kind of more in your face today than there is, there were a couple years ago. Um I feel like the, the main takeaway or the way to approach that is um having a good relationship with people on your team so that you're trustworthy. So if news is bad that's coming up, you can say it's bad, and take input on what to do with it. It's um, I think a misnomer that people think that they um are responsible if bad things have happened. And sure, sometimes they can be, but I feel like stuff is going to happen. There can be a big deal you were excited about that didn't pull through, or maybe an investment opportunity that looked like it was going to come through then the investor went dark or said no, or invested in a competitor. Like you can influence the world around you, but you can't control um nearly as much as you can influence. And when it comes to managing uncertainty, I feel like having frank conversations and being okay to deliver bad news. And having a culture where you talk more about "what are we going to do about it" than like lamenting the fact that something bad has happened, is really kind of key.
Because I feel like these things kind of follow from leadership at the top. So it's not really reasonable to expect, if I'm as a leader um super gregarious all the time and things are going well, and I won't say anything if something bad happens, to turn around and like expect your employees to speak up if something's broken or not working well, or if they're nervous about making a deadline. It just, it doesn't work. People mirror the um the behaviors that they see at the top. So that means if bad news happens or um, you just got to like bring it up and talk about it. I think one of the earlier ones that I did a bad job on was, I had an engineer that I was really excited who was going to join, and um I brought them in for a panel interview. Everyone talked to them. We gave them like an engineering coding take-home assignment. They did a great job, they talked about it. And then last minute came and they decided this is pre-COVID, they decided they didn't want to move to San Diego. So they, they said no. Um and I was like convinced that I was going to go convince the other person, so I didn't tell the team because I was still like, I was still fighting for it. And then their anticipated start date kind of came and I missed that, and um I didn't really, again maybe a friend at the dinner table, I didn't think through the perception from the team's standpoint. Like in my mind it's still, it's not a done deal. Um and then people, then like rumors start that way right? Because like you were super vocal, and all of a sudden silence. And then the outcome that you said was going to happen didn't. So I just feel like now when these kinds of things happen, you just mention it. You say, "Hey, unfortunately this happened, and um here's what I'm going to do about it. Like, um I could use your help on this. I could use your help on that." And having the culture where the team gets like mobilized to respond to it is much more powerful than like one lone leader out in the front who has to go fix everything because that's not going to work all the time. But if you can be frank and honest and bring the team with you, you can just accomplish so many more things in a day.
Jake Aaron Villarreal: Yeah, great. You know you talk about your company and you talk about a little bit about kind of the culture is kind of in there too as a leader. Um the book, Grow Up Fast: Lessons from an AI Startup here. If people want to get this, where would they go to get this?
Zachary Rattner: You can go to growupfastbook.com and there are links all over the place. Um it's on Amazon, on Barnes & Noble. We actually just released uh earlier this month an audiobook version. So if you like the [book] but you don't want to read it, um we uh, we have an audiobook out too.
Jake Aaron Villarreal: Is that on Audible or where's that located?
Zachary Rattner: It's on Spotify.
Jake Aaron Villarreal: Oh cool. That's great. Um where do you think AI is headed? I don't think you'll be able to get away from it. It's kind of going in multiple directions now. Um.
Zachary Rattner: So I think the first thing to say is no one, no one really knows all the details right. If you look at the pace of progress, um even 3, 6 months ago, the world has kind of totally changed. Um I can [give] an example I ran into is there's this new tool that's come out from a startup, I think they're based out of Poland, it's called ElevenLabs. And they do speech synthesis. So if you are for instance making an audiobook and your narrator um mispronounces a word. In the olden days of maybe December 2022, you'd go back in the studio, you record again. Um but with ElevenLabs what you can do is you can record somebody's voice, um and they have safeguards in place so you can't record other people, you can only record your own. You can clone it and then you can type and do text-to-speech and get phrases in your own voice that's come out. And there has been technology that's done text-to-speech for a really long time, but what's different now is that it is like almost out of the uncanny valley. I tried it on myself, and like I can convince myself that I'm listening to things that I never actually said. Which is super useful if you want to maybe narrate an audiobook in a language you don't speak, there's like great use cases for that.
But at the same time, I discover this tool, uh a week goes by, and I lose my debit card. I was uh taking something else out of my pocket, it fell, I got home, realized it was gone. So I'm calling Wells Fargo, and I'm trying to cancel it. And to get to talk to a person, the automated robot is saying, "Repeat after me, 'my voice is my password', please authenticate me." And like the world hasn't caught up to the advancements that's happening in AI yet, that like on the one [hand] this is possible. [On] the other hand, that's how you get into my checking account. It's kind of, kind of scary. Um please don't try. I probably shouldn't have put that on the air, but hopefully they patched it up by then.
Um but I feel like, what is AI going to do is a lot of the tedious parts of work are going to need-, they're going to go away. And that's because software will get better and people will be able to be much more productive, and it'll free up more time for the higher order parts of business, for being more strategic, for interpreting the data and trying to decide what to do with it. But I think the days of um like needing to crunch a bunch of numbers in a spreadsheet... I mean I even did it earlier today. I had um some AI performance numbers I was looking at and I put it in ChatGPT and I said, "tell me the aberrations." Like I just in plain English, and it's pulling through the data, it's spitting things out. Since it's early days, I'm double checking it myself. But I feel like it'll get to the point where you can just be way more productive. Um I prepared for this podcast by uh asking ChatGPT to summarize your last like 20, 30 episodes. Like I can listen to three or four, that was great, but I was like, "I want to get a sense of Jake over a long period." But um you can do things like this where you can just have way more output and no one has the time to do that amount of uh that amount of work. And I feel like those are the kinds of use cases that are going to grow up fast and kind of accelerate much quicker.
Um I don't think people will be irrelevant. I don't think you'll go and see a business where there's just no people, everything's being run by robots. But I do think you'll see um things like the first line support when you go and you chat in with a company. I think you'll see those chatbots like get better, and they'll be able to have people prefer that because they're getting an answer quickly and they don't have to wait in line. Um I think you'll see cases where designers will have an easier time making storyboards and coming up with ideas. Again I don't think you'll see AI generate a full feature-length film for, anytime soon. But like these early days where you want some sketches, you want some ideas for a new logo, those kinds of use cases I feel like where it's okay for it to make mistakes, it's okay for it to not be perfect, but it can make a lot and um kind of analyze large amounts of data that are very tedious for a person. I think those are the areas where we'll kind of see more and more advancement. So I'm actually pretty, pretty stoked. On there's a lot of fear out there and people talk a lot about AI taking things away, um but I feel like the opportunity that it's opening up is going to be so much bigger than the things it takes away that we'd be, we'd be fools for not wanting it.
Jake Aaron Villarreal: Yeah, I, I believe that. I also believe that the, the companies and the people that don't embrace it will be left behind. It's here, it's going to get better, it's going to get into all areas of work I'm sure. And uh it's going to be really fun to use other skills that you can learn and grow into versus what you do today that you know might take a lot of effort and mundane time but not necessarily be as creative or strategic. So definitely.
Zachary Rattner: Sorry, go ahead.
Jake Aaron Villarreal: I'm definitely going to um not just, just walk through this visually but with the audio and Spotify. I'm 100%, I'm an audiobook reader, listener. So um I can't wait to, to get into that in, in even more depth. Uh I want to kind of go a little bit back to your company. Sure. What in today's market is like the number one thing within your company you're trying to solve for or you're just you know focused on? Whether it's generating revenue, or more customers, or, or better product fit, or new innovation. What's, what keeps you up right now?
Zachary Rattner: Sure. So I think the, we exist in a larger ecosystem. Um market now is very different than it was in 2021. Um but I feel like the value prop of our product is kind of shining. Right, when things are going well and you don't care about your numbers as much because just invest in the stock market and you'll be better off in a couple weeks anyway, um maybe the need to run efficient processes is, is less, um is kind of less prevalent. Um but now I feel like folks do care about it. Um about top and bottom line, things like that. So things that, tools that make them able to do their jobs more efficiently. Also fuel costs, here in San Diego it's like over six bucks a gallon, so things that get people off the roads unless they have to and, and run better operations is, is going to be key.
Um so I think in terms of the things that we're focusing on now is we are adding on some new features. We're launching a couple new products in the, the next couple of months. And what we're focusing on is the general theme is allowing people to be more productive. Um I think there was a day that's come and passed where just having AI is like a cool bell and whistle and people like to use it. But um it's got to be real now, like it's got to drive economic value. The, the time that it takes to get to ROI can't be like 5, 10 years down the road, people kind of care about month-to-month, quarter-to-quarter. Um so we're looking at other areas. Now that we have this digital invent- that's kind of come in from the pre-move survey, what are other directions that we can go? Movers do a lot of things with that. They, they settle claims, they um, they sign estimates, there's a lot of different like workflows that kind of become enabled once you have that. So we're starting to look at like other directions that we can um that we can bring that into and bring more value to our clients.
Jake Aaron Villarreal: That's really good. Okay great. Well um Zach, if somebody wants to learn more about your company or learn more about your book, where should they go to find you, where should we go to find your company?
Zachary Rattner: Sure. So if you head to LinkedIn, that's probably where I'm the most active across all three: me, uh Yembo, book. Um so look up Zach Rattner, r-a-t-t-n-e-r, on LinkedIn. Um our company website's yembo.ai. But if you forget, yembo.com goes there too. So if you uh, you can learn more about the company, the product there. And then the book site is growupfastbook.com.
Jake Aaron Villarreal: Very cool. Zach, big shout out to you for joining us here. I know you're on a world, worldwide tour, right, getting your book out there and making sure that people are learning from some of the best. Um thanks so much for joining, thanks to our audience for listening. It means the world to us and I appreciate your time spending with us on this Friday afternoon. Um my name is Jake Aaron Villarreal, the host, and hope to hear from you and see you on the next episode. Ryan, thanks so much. Have a great weekend.
Zachary Rattner: Thanks Jake.
Jake Aaron Villarreal: Before we wrap up, I want to give a big shout out to all the entrepreneurs that have joined to make this podcast possible. And for all the listeners for listening, it means the world to me that you chose to spend your time with us today. I'm your host Jake Aaron Villarreal signing off for now, but can't wait to connect with you all soon on the next episode. Take care.
This show is sponsored by Match Relevant, a company that helps venture-backed startups find the best people in the market. And they do it in three simple steps. First, they sit down with founders to understand their story. Second, they tell their story into multiple candidate channels. And third, they schedule interviews within 48 hours. Find us at matchrelevant.com to learn more about how we do it.
.jpg)
.webp)
.webp)
.webp)
.webp)
