#529 AI Meets Talent: Federico Ramallo on Automating Tech Hiring Without Losing the Human Touch

[00:00:00] 

Mehmet: Hello, and welcome back to a new episode of the CTO show at me today. I'm very pleased joining me, and I think this is second time I'm hosting someone from Mexico. I have Federico Ramallo [00:01:00] Federico. I'm very pleased that you are with me here today. Why? Because first, you know, I, I tend to like to speak to founders, serial entrepreneurs.

People who are in the middle of what, uh, we call it, you know, the AI revolution, and you are a founder in that space. I don't like to steal the show from you. What I do usually is I keep it to my guest. So let us more about, you know, yourself, your journey and what you're currently up to, and then we're gonna have the discussion rolling from there.

So the floor is yours. 

Federico: Great. Uh. Thank you for having me today. Uh, it's an honor and a privilege to be in your show. My pleasure. So, a little bit about myself. I'm, I was born in Argentina, uh, from a Polish Irish family. Um. Um, it's, it's an, an interesting story. Um, I moved to Mexico almost 20 years ago and I currently live in Guadalajara.[00:02:00] 

Um, I am married to an amazing wife who's also my business partner. Um, I'm a father of a 10-year-old boy. Um, I hold a black belt karate, um, and, uh, I'm coaching on, on karate as well, uh, to my son and other kids. Um, nice. I am a software engineer and an engineer leader. Um, I've been coding for all my life and I'm, I'm passionate about building high performing engineering teams that deliver real user, uh, results.

Right? Um, I'm also a host of, um, of a podcast called Prevetted Podcast. Um. Where, uh, I, I talk with people that I admire, people that I find exceptional, uh, and interesting. Um, similar to, to what you're doing in, in your show, right. Um, I'm also, uh, founder of a few companies, which we can talk, [00:03:00] uh, in a moment. Sure.

And I'm also offering two books, uh, which we can, we can talk as well. Okay. 

Mehmet: Absolutely. We're gonna talk about that Federico, and thank you again for being here with me today. Uh, kind of traditional question, right? So whenever I have, uh, someone who founded companies with me, the first question I ask, what was the main problem you've seen that.

Let you build that company. Of course, you, you are a software engineer, so actually you build products, right? But what was like the exact use cases? And feel free to talk about what you're currently doing with Prevent AI and with your previous company Density Labs. I like to hear these stories. Federico.

Federico: Yes. I mean, um, I think that what happens is that, um. I built my, my, my company, uh, based on, uh, trying to [00:04:00] solve a problem. Um, so the, the company became more of, um, um, sub, sub, uh, sub consequent, um, of, of, of the problem that I was, I was trying to, to solve. Um, I was asked, uh, to build, uh, engineering teams, um, and.

That's kind of how I start Density Labs, um, which is a boutique engineering service firm and, uh, on Density Labs. We, we built, um, engineering teams in Mexico for companies in the us. Um, we started a company 10 years ago, uh, almost 10 years ago. And, um, it's interesting, the, the, um, outsource teams, um.

Dynamics. So the problem that I was trying to solve there was first having a, a reliable team, right? Mm-hmm. Then, um, I [00:05:00] found another issue that was we were not getting, uh, clear requirements from the product owners. So my solution for that was to build a design team that could help the product owners to provide more clear requirements, um, that way the amount of interruptions.

That the development team would have would be reduced, right. Uh, inter interruptions and dependencies. Right? We, we didn't need to go back and ask, uh, the product owners, uh, clarification. Right. Um, we could go, uh, and fly solo, if you like. Right? Um. And that allow us to, uh, first have more clear requirements on the design from, from the design team, because now we had mockups and, uh, um, what used to be a, uh, uh, a dragging meeting of, uh, having to go through the, uh, requirements document became a very, uh, fun meeting where [00:06:00] everybody wanted to participate because we added storytelling.

Into the conversation, right, and instead of having, having to go through a boring document. What we'd start doing is we start telling a story, right, with the mockups and showing how it worked, right? And that now everybody wanted to go there, even if they're not going, were not going to build that feature, right?

Because it was, it was fun. Um, and that allowed, allowed to, uh, facilitate a lot of the requirements, uh, and the communication with the team. Um, so we, we did that and. A again, there's a, uh, an interdependency right between the design team and the development team because now the development team could provide more clear, uh, more accurate estimations.

And I work with the team to make sure that, um. When we say it's done, it is actually done. It's not, it works on my [00:07:00] computer. Right. Which is usually the, the, the, the response from a junior, uh, en engineer. Right? Um, so I worked with that and then not only we got, uh, uh, accurate estimations, we were able to make sure that those features were shippable, right.

That they actually work on production. Right. Um. Which is another layer into this, uh, step of, of delivering features. Um, and in that process we also build a QA team, right? When you build a QA team, um, you, you have, you can have a QA team that could do manual test. Manual test and automated test, right? Um, and you can have the development team also writing automated tests, right?

So what happens is that if you're writing automated tests, if you're, if you're following TDD, you're doubling the effort for everything that you're building, right? So, [00:08:00] uh, then they start asking the questions, how much is it worth to, to do that? Right? Um. How much is it worth to do manual testing and how much is it worth to do automated testing?

Right. And working with the, uh, QA team to build a, a test plan before we're actually building the feature, because we could validate that test plan based on the value to the user. Right. And the effort to automate that, those, those tests. Right. Um, again, I'm trying to simplify, you know. 10, 20 years of practice into, into a, a single, you know, phrase.

But, um, that, that, uh, realization, uh, allow us to work with the engineering team to start thinking more on the QA side right from the beginning. Right. Um, on, you know, on on quality. The [00:09:00] earliest you find, uh, a quality issue, the cheaper it is to fix it. Right. Um, yeah. You know. If you see car recalls, right, uh, how much it cost to the companies to get the car, car cars back and fix whatever is the issue, right?

Um, whereas if you fix it, you know, earlier, earlier in the production line, you know, you save a lot of money. So, um, we start doing that and then, uh, we start working on, um, how much tasks. How much, how many tests should we do manually? How many tests should we automate? Because every time we automate, we free bandwidth from the Q 80 T.

Right. Um, because every time we do a release, they have to do the same set of tests. Right. Um, again, there are a lot of trade offs here. Um, but that concept was, was great because it allow us to optimize the [00:10:00] development team. Right. And. Making sure that what we were building would make sense, um, on the, uh, you know, uh, based on, on what the product owners were describing, based on what they, they, they talk with the users, right.

Um, so I, I've done that. Um, and one of the things, interesting thing that happens with the outsource teams or nearshore teams is that there is an in inherent power imbalance, right? Um, so as, as, as, as a leader of the team, uh, and working with, with the managers, uh, that, that, uh, are our clients, um, I work towards levering that play field as much as possible.

Right. Um, and, uh, the other. Uh, thing that I've seen happening on, on Nearshore team is that we are inherently shielded from the internal politics of the company. So we can [00:11:00] focus on, on, on, on the work, right? Um, so if, if we have parity, uh, in understanding the context, right, and we have the trust, the impact that, uh, the team can provide is way bigger, right?

We focus on impact and ownership. Right. And when we do that, um, and I push the envelope of the traditional role of the outsource team, right? Um, so we, we take more, more ownership, we become indistinguishable, uh, than a full-time employee. Right? Um, that, that's, that has been my, um. Secret sauce, if you like, of, of our, our success in, in density labs.

Right. Um, and then, uh, the other, uh, core, uh, realization that I've seen is, uh, I, [00:12:00] I've seen this happening on, on, on LATAM teams because that's where my experience come, comes from. Mm-hmm. But I've seen happening on, on teams, uh, on, on other countries as well, where, um. They, um, they, they are, they tend to over promise, and the way that I explain this is they over promise out of love, right?

Because they, they want to please, right? They want to say yes, 

Mehmet: right? 

Federico: And in particular in Latin America, what happens is if somebody invites you to a party, you say yes, right? Because if you say no. That's socially unacceptable because you are hurting somebody else's feeling, right? So even if you're not planning to go, you are, you should say yes.

Right? Uh, which, you know, causes an issue with, you know, the concept of RSVP in, in this type of meetings kind of breaks apart, right? Doesn't exist. [00:13:00] Um, so I work with my team to. Normalize, uh, you know, honest communication, right? Don't over promise. You know, let's promise what we can do. Let's, um, under promise and over deliver, right?

Because I know that that's the expectation that the manager is going to have, right? Um, and it's just a cultural difference. But, uh, these nuances make or break outsource things, right? Um, and then the other part that we, I work hard with, with, with the teams is deliver the bad news as early as possible, right?

Mm-hmm. Rather than holding them right? Because I've seen how holding them because of fear, because of, um, I don't want to, to get scolded. I'm trying to fix it. But, uh, you know, uh, before I say I say anything, um, that also [00:14:00] hurts the trust. That, that, that we have with the managers, right? Um, so delivering the bad news and then laying out the plan that we're doing to fix that, that issue, right?

Um, that way we provide the visibility, the plan, the north, this is where we're going. And, um, and that, that again has been, uh, uh, a differentiation of, of how I run these teams, right? And, um. The managers that work with, with us and then the labs appreciate it. 

Mehmet: Right. So Federico, how this evolve into pre-vetted ai?

Like how was the transition to there? Oh yeah. So, um, 

Federico: in that process of, of, um, building these teams, I had to deal with recruiting and screening. Technical people. Right. Um, so first I started with people that I know, people that I can trust, people that I can vouch for, [00:15:00] right? Uh, people that, uh, have been my friends or colleagues.

Uh, but you have so many friends, right? So eventually I had to build a recruiting and screening process, right? I am, um, very detail oriented and very um. Um, anal about, about quality. So, um, I worked towards building and probably overbuilt screening process. Uh, so my first goal was to have zero false positives and then, uh, I worked towards timing, right.

Um. So first quality, then timing. So if I, if I need to take a little bit longer to find candidates, I'd rather do that, right? So we build this, this, uh, screening process with humans. Um, so we, uh, we were able to, we were able to [00:16:00] screen consistently the best 2% of software engineers. Um, and we maintain a 98% success rate over the last six years, right?

So every, every candidate that we, we present. To, to the managers, to our clients. They has been successful in, in, in the, in the teams, right? And that means we set the right expectations. We work towards, uh, if there are, sometimes we find candidates that lack a skill or lack some experience. So we, we we're honest about that and we built a plan to improve, um, that.

That, that means it's a great opportunity for the candidate. It's a great opportunity for the manager because we offer a discounted rate during that transition process, and we give the opportunity for the candidate to prove themself. They appreciate that. Um, and that, that, you know, again, it's about building [00:17:00] trust, right?

So we build that process. It's been successful for, for, and, and proven, but, uh. The issue is we're spending 250 hours for every candidate represent, right? So again, it's o overbuilt, but with this idea of zero false positives, right? Um, so we do a, a few things. One is we, um, for every candidate that go through our process, we give them feedback on how to improve.

I think that's something, uh. Humane in, in this screening process. Um, I've seen, uh, screening process and recruiting processes that are more transactional. You know, um, I'm looking for that guy, you know, you're not there, you know, goodbye. And then they ghost candidates. So we took the commitment not to do that and to give feedback as much as possible, as much accurate as possible.

And the feedback usually is read this [00:18:00] book. Uh, uh, follow this, uh, this, um, uh, blog post, uh, you know, things like that, that are customized for that each candidate and apply again in a year. We want to see how much you progress. Um, some candidates follow that, some candidates don't, but at least on, on our side, we try to give them that guidance, right.

Um. So in that, uh, situation of 250 hours per candidate, um, we build, first, we productized our, uh, screening process. Um, so that way hiring managers can go to Preta ai and with three clicks they can see our pre-vetted talent pool. Our best 2% is available there for them to intervene higher. Um. So we did that, [00:19:00] that, that, that's the first step of Prevetted ai.

Um, and we're working on, uh, building an AI agent that can run the screening interviews right now. Nice. Um, having said that, I, I don't trust the AI agent yet. Right. So, uh, why? Well, there, there, there are a few factors. First is. The quality of the screening is as good as our humans, right? Uh, and then the second one is, is are the candidates, the really good candidates willing to be interviewed by an AI agent?

Right? I know that the, um, average, um, the mediocre candidates are going to, but again, we're aiming for the best 2%. So are they willing? That's a hypothesis that we're, we're trying to validate. Right? [00:20:00] So interesting. Um, my thinking is yes, I want to, to use the AI agent to run the screening, but I also want to have a human in the loop, uh, for screening.

Um. But supported with the platform so we can use the agent, uh, in a more passive role, just listening and, and, and scoring. Right? Um, or we can use the AI agent in a more active role, just running the interviews. Right? Um, now the other part for that is once we've, once we have that built for, for, based on our screening process, we can open up.

Uh, the customization for the hiring manager. So it, the idea is that the hiring managers will be able to go and define the screening steps that they want, and then we can execute with [00:21:00] agents or humans, depending on what the hiring manager prefer. Right. And then we can provide a standardized report on all the candidates that went through.

Right. So we're, we're, um, our first iteration is. We hold the truth, if you like. You know, it's, it's our process. It's our, um, it, uh, it's our screening. Um, with the agent. We are delegating on the agent a little bit, but again. I don't want to lower the quality, so I am doing baby steps. Uh, and then after that, we're going to allow hiring managers to run their own processes, right?

Um, which could provide different results. Uh, the output could vary, uh, uh, depending on the quality of the screening that the hiring manager set up, right? So we're going to provide templates based on, on our learnings, what work for us, so they can customize it [00:22:00] to their specific needs. 

Mehmet: Sure. Federico, out of curiosity, what you told me and you, you, you mentioned you work with companies in the us.

What I'm interested in, what type of companies are we talking here about startups, about enterprises? And the other thing is, uh, uh. Is it true? Of course I know the answer, but as an expert, is it true like now it's not like before, you know, especially before COVID, before 2020, telling someone, Hey, like, I'm gonna get you some especially technical resource remote.

No, no, no, no, no, no. I don't want to do this. I want someone sitting with me in the office. So. Two things. Uh, who are, you know, usually the customer profiles that you serve with the talent. And second, how have you seen, you know, the remote, uh, culture also facilitating that now it's not only with these kinds of companies, it can be any company.

So I want to hear your opinion and your, uh, view on that. 

Federico: Yeah, so, um, we work with [00:23:00] startups series A and above, and we also work with public companies. Also we have, uh, uh, enterprise and startups in nice in our client pool. Um, they have different needs. So the, the boutique and the personalization of the service, uh, has a, a big role on, on, on, on what they need.

Um, one of the issues that I've seen with startups in particularly, uh, young startups is that they expect the developers to be a, um, a. They, they open their, the expectation of the developers so much, right? So they expect the developer to be a little bit of product owner, a little bit of design, a little bit of qa, a little bit of DevOps, right?

Which some developers can do. Uh, but I usually, when, when we work with startups, I talk about, um, setting the right expectations for [00:24:00] developers, um, because it's not the same and. Kind of, uh, going to what you're talking about, the expectation of have having somebody physically in the same location, um, for a startup that is all over the place?

No, no. Uh, formal process set up trying to build something fast. Um, usually having a local developer is better because they can just tap their shoulders, sit down together. Uh, they can do this more, uh, chaotic. Uh, a creative process, if you like. Right? Um, but when you, um, but when you have a remote developer.

Then, uh, having a more formal process, having, uh, more, uh, clear requirements, having a, a better defined role, a narrow, narrow role for the developer. Uh, it's better because, uh, re [00:25:00] remote, whether it's in the US or outside the us, um, there are, there are communication issues, uh, that could affect the out outcome of, of, of the developer.

So, um. Having, having, uh, un unless you have a remote first organization right. Where, where that it doesn't become, uh, a big issue. Right. Um, but I, I've seen, uh, bad practices in that sense where, uh, the developers are working on, on all hours trying to, to get to the next demo. Right. Uh, which again is startup life.

I understand. Right. Um, right. But one of the differentiations that I've seen in, in this more a cultural difference between the US and Latin America is Latin America values more family time. Having a work-life balance is much more important for them. [00:26:00] Right? So, um, having a, um, a Latin developer in a, in a startup.

Becomes an issue between, uh, time balance, that a work-life balance and communication, right? 

Mehmet: Yeah. Now one of the things that I heard even here on my show, uh, that CTOs usually they worry about it, is long-term retention. Uh, especially with remote teams. Uh. How do you make sure that the talents that they are prevented, uh, for these companies, uh, are really ones who.

Would be able to build the loyalty and the trust, especially for an organization that they might be doing this for the first time. They're worrying. Okay, maybe they will, we will hire them for six months, then they will find something else. They will leave [00:27:00] us. So how do you manage this Federico? Or how do you advise actually the talents to, to, to do so?

Because at the end of the day, if, if you recommend someone to a company and they do this, it's not, it's gonna hurt you also as well. Right? So you need to make sure that the, the, the, the, the talents that you are presenting to these companies, they're gonna be thinking long term. So how do you manage this?

Federico: Our average tenure is six years. 

Mehmet: Wow. 

Federico: And we've been in building for, uh, we, we, we have been in business for 10. So yes, it's very high. Um, it, it exceeds the average tenure in Silicon Valley, for instance. Um, it's, it's a lot of work to, to to, to have that high level of retention. Um, but um. We [00:28:00] work a lot on ownership.

We work a lot on building a career path. We try to, um, either, uh, read between the lines or, um, or ask the, the, the, the CTO or the manager, what are their future needs. Um. So that way our team can work, uh, as part of their career path on learning that new technology or a new framework or whatever it is, it's going to be on the roadmap.

Sometimes they don't know. So we, we build a, a career PI path based on our assumptions. What, what are things we see are they are going to need in the future? Um, and that sense of ownership and growth. It's the key to keep the team happy. Right? And then on the other hand, we, uh, we, we try to, we, we work towards a no [00:29:00] handholding, uh, position, right?

We don't want, uh, a developer that you need to kind of, you know. Uh, spoonfed, right? You want somebody that can of course, get in and, and work, right? So, um, we work towards, um, setting that, and we do the same thing with, with the career path. Again, we, we try to reduce the, uh, the bandwidth that, that the CT o requires, uh, to manage the team because we, we proactively work towards that and.

Also provide them clar, uh, um, provide them, uh, visibility of what we're doing right now. On the other hand, uh, we usually have a backfill ready. So in 24 hours, 48 hours, you can have a, another candidate working in case in the worst case scenario, right? Um, so we plan for the best. Uh, plans for the worst work to the best, right?

Um, yeah. And then, [00:30:00] uh, the other thing that, um, that we do is, um, manage expectations. Understanding what are the expectations of the manager, so that way we make sure that the developer or the QA or the senior can fulfill those. Um. Those expectations. Right. And if they don't, then I, I mean, I believe that my goal is to be able to be the first one that raise the flag and say, Hey, we have this issue.

And not only we have, we have this issue, but also, uh, this is the, this is the action plan that we have in order to fix that issue situation. So rather than coming with problems, I come with. Situations and an action plan. 

Mehmet: Right. 

Federico: Uh, and last, um, I can give you, um, sure. Uh, an example of one thing we did is, uh, we built a trainee program.

Um, uh. Usually we start with senior people that join the team, [00:31:00] and then we work towards, uh, trainees. Right? Um, so what we do with the trainees is, uh, we train them ourself on our dime, um, to make sure that we know them. Uh, we check values, we check behavior, we check, uh, because you can interview them, but usually the knowledge they have is very, uh, limited.

So we. I, I focus more on behavior and ethics and, uh, integrity, right? Uh, because I can teach the technical stuff. I can teach to, to, you know, any, again, people that having screens so good enough candidates, right? Uh, so I want to, to do the, the, the, uh, other part of, of, of the human, uh, human profit, right? So.

We train them. Uh, then, uh, 50% [00:32:00] of, of that, uh, training team, uh, are going to be, um, elected to the best. 50% is going to be elected to, to participate, uh, on, on the team, um, the hiring managers. Usually wants to participate in that, in, in that process. But our intention is everybody that goes through our screening process and everybody that goes through these initial, um, um, training, screening, uh, everybody has to budge for them.

So, um. Basically we say this is, this is the people that we feel are our best. So they work really hard to become the best, you know, half right, the, the, the half that are going to be selected. And it's a learning opportunity for them. Um, we pay them to, to, to do that so that that way they can have a focus on, on, on the training.

So it's a learning opportunities, uh, income opportunity [00:33:00] for the candidates. It's a great way for us to know them. Then after that, we work towards a three year plan with a soft commitment with the manager. If a candidate is not performing, then it is not performing. So in those three years, we had a hundred percent retention rate, right on both sides.

So it's, the manager is happy with the work, the, uh, the trainees are happy working on, on the, on the project. Right. Um. And then we, we planned, um, raises, uh, salary increases throughout those three years. So that way they don't have this uncertainty of, am I going to get a raise or not? Right? But also, we work with the manager on setting expectations for every EEE, every, uh, you know, e every stage.

Again, with a soft commitment. If, if the person is not doing, uh, is not delivering the expectation, then we can move a little bit the [00:34:00] dates. But the intention is to have a, a plan that that allows, um, that, that, that allows to, to have a high retention rate. And then after those three years. And we've done this pro process multiple times, right?

With a hundred percent retention rate. Um, after those three years, then they usually stay on the project. We had a hundred percent, um, hiding after the, after those three years, which basically means they continue working on the project, right? Because they're, they're, they're performing. Um, and we've seen six years of, of, of, um.

Of tenure. Right. And the the reason for that is because they appreciate the opportunity and the growth. They have growth throughout, throughout those six years. Right. That is associated with new responsibilities that they take. 

Mehmet: Right. 

Federico: And the managers [00:35:00] appreciate the, the work that that, that each of these people are doing because they're pro 

Mehmet: providing real value to the team.

Cool. Now, you mentioned that the introduction about the book or two books? Yes. Tell me, tell me more. 

Federico: Yeah, so, um, I'm writing two books. One, I'm the co-author, uh, of Open the Valve, uh, the New Path to Creative Play. Um, I'm offering the, the Spanish version. Uh, okay. Which it's,

uh, so this book is about creativity, right? And um, the. It's an interesting, uh, journey. It's a non-linear, interesting journey about how to, uh, become creative and how to embrace creativity. Right. Uh, so it's inspirational [00:36:00] and I'm working on, on the Spanish version, uh, and it's a back and forth, so I I, I've been adding contributions on the English as well, so it's, it's a very interesting.

Dynamic. Right. Um, and my coauthor, John Clemson, he's an amazing, uh, author, an amazing guy, and, and he's, he's, uh, he invited me to participate in this project. Um, and that the other book that I'm, I'm writing is about leveraging outsource teams successfully, uh, which the practices that I've been doing for the last 10, 20 years, and, um.

How managers can advance their careers by leveraging these outsource teams. Right. Nice. What about the podcast? So the podcast, um, I, I talk a little bit about the, the, the book about creativity and the podcast, um, became. First, it's a way for me to [00:37:00] connect with people in a much more meaningful way. Like, you know, this conversation we're having, but then I realized that I am becoming the creative conduit for my guests so they can share their stories.

Um, they can brighten up you, you can see when they, uh, start talking. And you probably experienced this as well, when. Love you get into the flow of the conversation and it's hard to stop it because, you know, everybody's on the flow and it's such an amazing experience. Right. Um, so I, I found that there is a connection between the book about creativity and my work on the podcast.

Mehmet: I can relate to that. So what's the profile usually of the people you host now on the podcast? Federico. So, I mean, are they, are they CTOs, are they founders? Like what, what kind of guests you have? So, I, I had a 

Federico: [00:38:00] very interesting guest. Uh, everybody has an interesting story. Um, of course. So I had, um, book authors.

I had, um. Uh, individual contributors, uh, product owners. We talk about product. Um, I've been doing a series of episodes about that. Um, I had CEOs, I have, uh, engineers, former NASA engineers, uh, where we talk about, uh. Uh, putting radio telescopes on the dark side of the moon. Right. Nice. You know, it's very technical, but you know, it's, it's a very interesting conversation, uh, similar to the VO telescope that, um, collapsed, uh, a few years ago.

Uh, you know, where, uh, uh, there was a James Bond scene there, right where Just die. Right. Uh, spoiler alert. It's a, [00:39:00] it's a movie from the eighties, I think. But anyway, uh, we, uh, I thi I think Pierce Broman is, is the, uh, oh oh seven. Uh, so anyway, um, we, we, we talk, um, about life sometimes. Um, uh, I had, um. I had, uh, guests that, uh, uh, for me it becomes a way to kind of embrace life because I had guests that, uh, their kids showed up, you know?

So, uh, you know, instead of kind of, uh, I embrace interruption in a way, right? Because I believe that we are one. Right. Uh, as, uh, family businessman, you know, uh, developers, you know, we, we are, we are one and, and, uh, and, and everything is, is is interconnected. So, uh, I think it's, it's important to embrace this idea of, uh, that, [00:40:00] uh.

You know, we're also having a family. We also have, uh, you know, uh, uh, our kids are part, are part of our life. Uh, our dogs are part of our lives or cats, whatever it is, right? Um, so I try to embrace that, right? So, um, yeah, different people talking about different things. Um, I had CEOs as well. Uh, I, I think I already mentioned that.

Um, and one of the things that we've been talking a lot with. With, with my guest. It's about, uh, this idea of, uh, running business for profit business with a social component. Um, we kind of gravitated towards that. Um, again, I know that there is a connection, but I know the causality, right? Yeah, because I, I, I, I have that, that idea, and I, I found that a lot of the guests participate.

My, in my podcast bring those, uh, ideas as [00:41:00] well. So maybe it's a serendipity thing where we kind of connect for that. Um, yes. And um, for me, the podcast becomes a way to do simples acts of kindness, to, uh, provide the platform to the guests to, to share their story, um, to help them be more creative. Um, and, and Bright through the episode, and it's amazing.

Yeah. 

Mehmet: It's, uh, I would say it's very similar to this podcast. Uh, how long, how long have you been doing this Federico? Like how, how, how long have you been doing the podcast? 

Federico: So I've been doing the podcast since 2023. I started, oh, same 

Mehmet: time I started. Oh, really? Amazing. Yes. Yeah. Uh, the reason I ask you why, because almost you know what all, what you mentioned, of course.

[00:42:00] Maybe we didn't have the same guest. Like, funny enough, maybe, you know, I have you now, probably you will have someone like me later. Uh, but exactly the same concept. I had CEOs, CTOs, CMOs, uh, uh, I had people who talks about everything, technology and non-technology, but we try to relate it to our daily life.

And, you know, technology is part of our daily life. So I, I, I like, you know, this way of doing it, uh, this way. And, you know, I, I'm, I'm fascinated already, uh, by the power of doing this, as you mentioned, is about creating impact. And it becomes by itself, your launch pad, right? So you figure out after doing it for a long time, oh, oh my God.

Like probably I didn't know I'm doing this, but people are benefiting out of a content that I'm creating, which is. For me, it's fascinating. It, I can't, I can't describe it in, in a proper way, so, uh, I'm happy. Like also you found a passion for you [00:43:00] also other, of course what you're doing very well with, uh, with your companies.

Now, speaking back to to, to that, uh, Federico now things are changing fast. Ai you mentioned, you touched on it a little bit. How are you seeing the future? Of this business from your perspective? Because, you know, and the reason I'm asking you people are asking are, are saying, yeah, AI will write code, AI will do this, AI will do that.

From your perspective, what's, how do you see things going within your business when it comes to providing, uh, you know, pre-vetted teams? So 

Federico: there are a few things to, to talk about here. It's, uh. We can talk about AI in the cons, in the, in the terms of screening people. Um, there is this concept of if the employer is using ai, then as, as a candidate, I can use [00:44:00] ai.

Um, I draw the line on, um, on honesty and integrity. So I've seen people using AI to cheat. Uh, a candidate using AI to cheat and, uh, present themself as something that they're not so covered use of ai. I think that's, uh, a, a, a big no-no, from my point of view. Right? Because, uh, it's, it's, it's using to be, uh, it, it's deceiving, right?

Um, now on the other, on the other hand, um, being able to talk honestly about how a, a candidate is leveraging ai. To, uh, provide answers to, to do work. I think that's, that's something healthy. Um, but it should be out in the open and, and over the table. Right. Um, now I also see the, [00:45:00] the other, the other point of view of candidates saying, well, we get automatic rejections.

Um. And, and that's it. So I understand the frustration of, uh, people, candidates feeling that AI is being used to, um, provide a more, uh, personalized rejection letter, right? Uh, and not talking to humans, right? Um, so I, I, I also see that that's why on, I mentioned about, um. You know, maybe our AI agents doing screening would not apply for everybody.

Um, and we would need a human to step up into the conversation. Uh, that's why I I do the, um, the feedback because I believe that providing a genetic re rejection letter, uh, does not provide value for the candidate. Um, and if we can leverage AI to hyper personalize that feedback, the better. [00:46:00] Right? But. For the key differentiation for me is, um, follow up, right?

Uh, are you willing to come back in a year, six months after you, you read these two, three books and now you know more, um, or, or not, right? Most candidates don't follow up, right. And that's a missed opportunity that they have, uh, which is not ai, it's on the human side, right? Um, okay, so I talk a little bit about, about that.

Um. Is AI going to replace engineers or can companies replace, uh, engineering teams with ai? Um, I believe not. Uh, there's a lot of hype on that. It sounds great. Uh, for non-technical people, um, on the business side, sounds great. But then, uh, when you start, um. Implementing things AI does, doesn't have this level of abstractions that humans have.

[00:47:00] Um, so I believe that, um, engineers using AI or how knowing how to leverage AI are going to replace engineers that don't. Right? So that, that's, that's, uh, I, I think that's going to happen. And then, uh. For, and particularly for companies that are, uh, legacy, that have, you know, enterprise have big, big, uh, code bases.

The use of ai, uh, you know, replacing engineers with AI becomes much more dangerous because there's so many moving parts that, uh, and trade offs that you need to consider when you're building this type of software that is, is not feasible. Right. Um. For prototyping for product owners using AI to convey the, their ideas for non-technical people using AI to build something that otherwise they couldn't.

I think that's a great use for the tool. Um, I'm a Ruben Rails. Uh, I love Ruben [00:48:00] Rails, but, uh, if I have to, and I learn bash because it's a hard language and, um, and I love it because of the challenge, but if you, if I have to write something in Python. Now I can use AI to help me write a script in Python.

Right? Right. So now as a software engineer, I don't need to know all the languages I have, I have more, more tools, uh, available for me. So I can write in Ruby, give it to an an LLM, and. Translate that into Python or, um, you know, kind of stumble around, uh, and build something that, that needs to be done without having to learn the whole language.

Right? So there are benefit for, for LLM on that, on that, on that side. Um, but yeah, I don't think it's going to replace humans yet. Yet. 

Mehmet: Yeah, exactly. Yet. Uh, uh, as we are almost, you know, coming to the end of this conversation, which is fantastic by the way, Federico, one question that actually I wanted to ask before, but [00:49:00] now I have felt it's the right time to ask you.

You've done and you've built this fantastic business, and you've been almost 10 years now, and your sweet spot is between Latin America and the us, right? Yes. Now, if someone's sitting and listening to us, I don't know. Between two geographies, which might be on the same time zone because I know like, uh, latam and us almost same time zone.

Of course you have, I don't think even the language is a problem, but I mean, at least they are working on the same time zone. If someone now, I don't know, maybe sitting in Vietnam and thinking to provide this to, I don't know, maybe a geography, which is time zone perspective, if they want to do it the right way, what are like, you'd say two or three things you advise people.

To build this in the right way, unless you tell me MeMed. No, I will not share because I'm planning to expand and make this business global. So what you say? 

Federico: Well, I mean the, uh, first [00:50:00] the idea of Prevet was to expand globally, uh, which now we can do. Um, and, but then lab has been mostly focused in Mexico. Um, uh, so.

There is that, that intention. So if, if there's any candidate that want to apply on prevented AI from Vietnam, they, they, they could, um, could Now, um, one of the, the, um, issues that I see with working in a different time zone is, um, think of the experience as a manager. I give you, I'm the, I'm the manager. I give you the requirements.

Please build this. I go to sleep the next morning I see what's, you know, I see the results. Right? Right. Yeah. So what happens if that person, uh, that developer work in the wrong direction? You, you learn that, uh, as a manager, you [00:51:00] learn that the next day, right? So there is no feedback loop. Uh, uh, so the feedback loop is eight hours or one day, right?

Whereas if you're in the same time zone, you can have feedback loops every hour, you know, or as frequently as as you need because you're working on the same time zone, right? So that's, that's a big benefit. So you can find a, uh, different, different rates and you can say, well, uh, it's going to be cheaper in Vietnam.

Um, but. I, if I, if I have the person working in the wrong direction for eight hours, then the higher rate becomes much more appealing. Right? Um, so there's that, that, that, that situation. Um, then, uh, I think that some of the cultural, um. Differences that I mentioned. Uh, there are nuances in, in, in, in, you know, Pakistan, [00:52:00] Vietnam, you know, uh, there are nuances.

So I, I don't know those countries well enough, but, um, I would suggest some, if somebody wants to do that, they should, uh, go to the. Work for a while there, understand the culture, right. And then come back and be able to become this cultural bridge. Right. And set the expectation on both sides. Right now, the time zone difference could be solved by having the people working at night.

Uh, but I, I don't know how much, uh, they can sustain that in long term. Right. Um. Especially 

Mehmet: back to your point of the life work, uh, balance. 

Federico: Right? Right. Uh, I believe that in Vietnam and you know, other countries, they're going to appreciate work-life balance more than in the us. Um, so probably that's going to be an issue.

Um, and one of the things that I've done, [00:53:00] uh, and one of the reasons I I chose Mexico is, um. We can be in from Guadalajara to San Jose in three, 3.8 hours. 

Mehmet: Yeah. 

Federico: Right. Uh, so we could go, uh, and, and we, we've done this before COVID, we would go four or five times a year, just work locally and. Integrate with the people, right?

So instead of being the product owner that it's, you know, uh, uh, an avatar on a, on a screen is the product owner that, uh, she beat us on mini golf Right. Uh, a few weeks ago. Nice. Right. So that personal relationship, solds or, you know, uh, facilitates the communication much more. Right. Um, so that's, um. That benefit, we can do it from, from Mexico.

Uh, [00:54:00] when I was working from Argentina, I had 17 hours flight to the US right? So it wasn't as easy for me to to do, right? So if, if you are in, in Vietnam for instance, you have a 20 plus hour flight, right? Uh, plus all the time zone shift that adds, um. Um, um, jet lag, right. Um, of course. So, uh, you have that challenge.

And then the last one that, um, I found that, uh, benefit from Mexico, uh, is, um, we, we had clients shipping us, um, test devices, 

Mehmet: uhhuh, 

Federico: and the shipping cost was 20 bucks, 50 bucks. Of course, 

Mehmet: you know? Of course, yeah. Very affordable. 

Federico: Yeah. And. To, to that point, I actually drove, I, I've done this like 10 times and I love it.

It's, uh, a little bit more expensive to travel, but, uh, I drove from Guadalajara [00:55:00] to San Jose to San Francisco, uh, 10 times already. And, uh, you know, I, I actually, instead of shipping the devices, I will bring them with me to the, to the office, right. Uh, because I drove there so I could put them in my trunk and, and drive back.

That's something you cannot do, um, with any other countries. Right? True. Um, and if you ship it to, um, Vietnam, it's going to cost thousands of dollars. Right. And the risk that that device could get lost in the process. Right. 

Mehmet: Absolutely. Absolutely. Yeah. And, and this is your sweet spot, uh, Federico. And, uh, uh, you know, I appreciate what you do and, you know, all, you know, building this business, you know, just being 10 years doing the same thing as they say in French, like, uh, congratulation on, on, on this because really, uh.

Especially your business, and I speak with all by the way, of even on the show, I spoke with a [00:56:00] lot of people who are like, in similar domains, not doing what you do, but similar like recruitment offshore, you know, building augmented teams and all this. But I mean, this, this kind of business needs, uh, patients because, uh, 'cause you need to understand the other side.

Finally, question Federico for you, where people can get in touch, where they can find about you and about Ted ai. 

Federico: Yeah, so, um, they can find me on Density labs.io, um, prevetted ai, uh, that's the, the Prevetted platform. Sure. And pre Prevetted AI slash podcast. That's where we're hosting the podcast and they can, they can see, uh, our episodes there.

Mehmet: Great. Every single link. And you know this because you are a host, so every single link that Federico just mentioned, it'll be available on in the show notes. If you're listening on this on your favorite podcasting [00:57:00] app, you'll find it there. If you're watching this on YouTube, uh, you'll find in the description.

Federico, I can't thank you enough. Really, I appreciate the time. Uh, it's a nice experience for me to learn about this. Uh. Business and this niche that, uh, you found and you succeeded in of, you know, uh, getting teams For US company from Latin America, it's a big impact. I can feel this because also like it's getting the guys who are like.

Talented in Latin America to also know how things are done in in, in the us. So this is a big, big, big impact, what you are doing with the podcast, the two books who, which are coming I think very soon. So congratulations also on that. And this is how I add my episodes. This is for the audience. If you just discovered us by.

Thank you for passing by. If you liked what you heard, what you saw. I really appreciate if you can give me a favor and subscribe and share it with your friend and colleagues like Federico is doing. We are trying to do an impact. We're [00:58:00] trying to reach as much as people as possible. We're trying to touch people lives because maybe Federico said something today, which.

Park something in, in, in you. Maybe you'll find an opportunity, maybe you will connect with Federa and you know, he, he will be able to help you in some way, or maybe you will be able to help Federa in some way. It happened on this podcast before. That's why I'm saying share it with your friends and colleagues and if you are one of the people who already do this and send me their feedback, they send me their questions, they send me their, you know, comments.

All comments, whether positive or negative or appreciated. I read all of them really guys. So keep them coming. And finally, thank you for making the podcast this year. I know you get bored of me repeating this again and again, staying in the top 200. Apple Podcast charts across multiple countries. So of course we keep, we keep changing this time for Mexico, let's say Federico because of you.

I'm hoping, like when I get some, you know, a top 200 chart in Mexico, [00:59:00] I did it in Canada, I did it in Argentina, I did it in Bolivia, I did it couple of, uh, latam countries. So maybe this time for Mexico. And of course I'm waiting my US friends as well. I didn't enter your top 200 podcast chart, but I'm hoping I will do it soon.

And finally, uh. In September last month from nowhere to next was out. This is my book, which is, it's not only about me, it's about all the thoughts that we discussed on the podcast in the past three years. So you can grab it from Amazon. Thank you very much and stay tuned for a new episode very soon.

Thank you. Bye-bye.