Hey there and welcome back to lead the Team.

Most people are just trying to keep up with AI, but Lucas Mendez is actually building the workforce behind it.

He's the Stanford trained engineer ex Goldman and Bain, who is CEO and founder of Revello, powering both the teams who build AI and the companies racing to adopt it.

Today you'll discover where the future of work is really headed.

Lucas, welcome to lead the team, sir.

Thank you, Ben.

It's a privilege to be here.

Looking forward to this conversation.

Only took me a few times to get that intro right.

Woohoo.

So you, you started by placing people in jobs years ago and now you're actually helping machines do those jobs.

Tell us about that.

All right, that's, that's 12 years in five minutes.

We, we have now.

Okay, yeah, great.

After I sold my first startup to a private equity fund, I went to business school at Stanford and I started looking at several different areas of business that I was definitely willing to explore and things that I wanted to start new businesses in.

And the world of work was really something that attracted me because it is an unsolved problem finding jobs or looking to find talent.

It's so messy, it's so many holes you can fall into.

And yeah, I just thought that technology could lend a big hand at improving that market.

And the first iteration of Ravelo was actually what you just described.

We were helping people find jobs.

Specifically we were focused on engineers, technology professionals who were in very high demand, who had very specific skills and who just didn't fit with the average job discovery process.

Right.

So we created what essentially was a glorified job board.

Right.

So it was a talent marketplace where everyone had to take tests and assessments to join that marketplace.

So that gave companies access to pre vetted talent, which was great.

But we were doing that for a very specific part of the world.

We were doing that for Latin America.

And the reason for that is that I'm from Brazil, I was born in Brazil.

And even though I, I worked at those U.S. companies and went to school in the U.S. i still had the connection with the, with the region.

Right?

Yeah.

And for those listening, if you don't know this, Brazil is a giant, giant market.

Yeah.

Brazil is the 10th largest economy in the world, is the second largest, largest user base of Facebook and you know, Twitter as well.

So basically it's a very hyper connected country and there's a very large VC market there.

Anyways, we started the company there, the company grew, we raised some money for Velo, then we hit a wall.

We kind of hit a wall in around 2019, where we just saw that the addressable market for what we were building was just not big enough for us to hit a home run.

Right.

So that was clear.

And then Covid hit and somehow things got even worse because From March to April 2020, our revenues dropped by about 80%.

And we thought, we're done, we're done.

That's it.

By stroke of luck, we had recently raised a round of capital, so we had a pretty full war chest.

Right.

And that kind of allowed us to test different iterations of the business and try to get a feel for what could be next.

Right.

And then the LATTER Part of 2020, that, if you guys remember, this is when remote work kind of became a thing.

Like, everybody started working from home, Teams started being built.

Like, yeah, I mean, people started building teams, internationally distributed teams.

And we started getting inbound demand from U.S. companies who were saying, look, my tech team is now working from home.

I might as well hire the best talent from Latin America because I'll be paying significantly less for talent that is just as good, as nearly as good as the talent I can find in Silicon Valley or New York.

And we started reacting to that demand and all of a sudden our market just blew up in terms of size.

The company grew seven or eight times in four years.

And yeah, and then, you know, that, that gave us a, the, the biggest second win that you can imagine.

And that's that.

That's the second part of the story.

Now, to your question, how do we then get into LLM training?

How does.

Online like you're posting jobs for engineers?

It's not AI related.

I mean, at that point there was no AI.

No.

And then two years ago, we started getting inbound again from companies who now are in the AI space saying, hey, our models are getting better.

And you know what?

One of the biggest use cases of LLMs then and today is to help engineers write code.

But for them to be good at that, just regular LLM training doesn't cut it.

They need something called post training, which is feeding the model with extra specialized data produced by humans.

That lets the model become even better at specific skills.

And they said, not just any data.

You'Re saying specific technical data.

That's good data.

Not just anything.

High quality data, specifically code.

If you want your model to be good at writing code, you need to feed it amazing code.

Yeah.

And by the way, that applies to any leader out there who's considering AI for their business.

And we're going to dive into more of your world.

On this.

But it's not enough to just be the leader using AI.

You got to know what is going into that black box machine before you're basing your strategies, your technical expertise, anything in that.

And that's where you guys came in here now.

That's it.

That's it.

And all of a sudden, you know, to generate that data, these companies need hundreds, thousands of software engineers of the highest caliber that are producing that code and feeding the models.

And guess what?

Over the years, Revel had built this network of over 400,000 pre vetted engineers across Latin America.

And we said, you know what, we can build a platform on top of this.

And then over the next following two years, we built a technology on top of that network to allow us to collect that data, curate that data, to, to do the quality assessment on that data.

And today we offer a lot more than just the people.

We offer almost a consulting like services service to hyperscalers who are building their LLMs.

So yeah, that's it.

10 years and 10 minutes.

So you are building a global army of AI engineers feeding quality LLM information and data to basically fuel AI globally.

That's it.

I mean, that's correct.

And we look at the backbone, baby.

That's the backbone.

It's the backbone and it works in two different ways, man.

I think there's like, there's the hyperscalers and AI labs who are building those models and they need that talent.

So it's the backbone for that on the one side, but also every company and their neighbor want to implement AI.

Every CEO out there is giving their technical teams mandates, guys, let's use more AI.

And the way to do that is by leveraging tech talent to build AI solutions internally.

Right?

And that's where we come in.

So we do both the, the AI development and the AI implementation.

So that's, that's what we like to call the backbone for the edge of AI.

Yeah, so they work.

So they work.

The people that you were trying to get jobs with other companies, now they work for you.

And you're able to help them funnel into these larger organizations in a very specific way to provide this great data that they need to fuel their internal operations with AI.

So in other words, if you're a big company leader, you don't need to be thinking about ChatGPT.

You got to be thinking about how you're going to use quality LLM and AI solutions for your company.

And you have this backbone of engineers that come in and understand the business.

I don't want to Put words in your mouth.

But I'm like, really getting this business idea, this concept, and it seems so vital.

It is.

I mean, whether you're building AI features within your product or you are creating the data pipelines within your business to use to make more informed decisions, then you're using AI.

You're like, if you're not an LLM foundational model company, if you're not OpenAI, if you're not anthropic, if you're not Meta or Amazon or Google or etc.

If you're not one of those guys, then you have no business building your own model.

You're basically plugging that model into more structured solutions that can drive value.

And for that, you need talent.

The way I like to think about it, Ben, is investors talk a lot about Nvidia, and Nvidia is a great company.

I'm a huge fan of what they're doing because they are the capex of AI.

Anyone who is investing in AI needs to buy the building blocks.

And the capex for that is Nvidia is GPUs.

The opex of AI is tech talent.

So the way that we like to think of ourselves is we're the providers that will essentially become the opex of AI.

Wow.

So what are the questions?

So leaders now they're like, well, well, Lucas, dang it.

I just thought I was gonna let all my employees play around with chat, GPT and Gemini all day, and they're gonna solve our customer solutions with that.

And you're, and you're saying, and they're probably listening to this and they're like, okay, that might not be the best way forward.

The questions.

What are like the foundational questions that they should be asking themselves to move forward from today into the future with AI?

I think it's a question about driving value, Ben.

I think the main question is you can talk about AI, but you can also talk about technology, about mobile or social.

Ten years ago, that was the new thing.

Ten years ago, Right.

And the question is not how do you.

The question is not asking your cto, hey, get me one of these.

It's more like, how can we think as business leader about ways to leverage value to our end client using that?

That's the key question.

And that can be a faster service delivery.

That could be an improved user experience, that could be more flexibility in terms of how you serve them.

But it's not about the technology itself.

It's about how much value you unlock.

What happens to global workforces when AI starts doing what Your clients are hiring people for.

That's a really, really good question.

Look, I'm an engineer and I've gone through many cycles of technology where people say, okay, now engineers are done, that's it, everyone will become an engineer and they're all going to lose their jobs.

I don't think that that is happening anytime soon.

Not because the AI is not good.

AI is great, it's writing great code.

But the job of engineers go so much further than that.

Engineers around the world, they are problem solvers basically.

They need the technology to develop solutions that solve problems.

If the technology is punched cards, as it was in the 70s and 80s, that sucks.

If the technology is then mobile apps and web apps as it was in the 90s and early 2000s, that's better.

If the technology is AI enabled tools, that's even better.

So my point of view is that engineers as problem solvers are going to be in so much higher demand over the next 10 to 20 years than they were in the past 20 years.

And it's a good thing then, then every, that everyone will become an engineer because everyone will be able to write those solutions and, and, and solve problems.

So what I hear in that is, hey everybody, the bad news is your job today, as it exists today as an engineer, will be replaced by AI.

But the opportunity is what you can do with it tomorrow, the next day is, is to stay ahead of it.

So because your job today will go away, probably all of our jobs to some degree, but it's how we ride that wave into the future.

How do we keep our edge, how do we keep developing?

So we sort of stay ahead and stay of AI and harness it as a tool, you know, and that's it.

That's it.

And I think for leaders specifically, who I think are a large part of the audience listening to us right here, it's about pushing the team to think differently.

Just using the same mental models that we've used for the past 10 years in this new age is not going to cut it.

People are going to be much more powerful agents in solving their own problems.

So you need to screen for that.

You need to build teams who actually can leverage these technologies and you need to push them in the direction of using them and unlocking more value.

So I think that the role of a leader is going to change dramatically as well during that cycle.

Let's talk about one of your favorite topics which we know is luck.

You already mentioned luck once.

I know, before we got on here, our chit chat, you love talking about luck.

Let's talk about, I want to talk about luck in your career, but I also want to talk about luck and AI.

What is the role that luck plays in AI and how can leaders benefit from that?

All right, that's a good one.

So for me, luck is being in the right place at the right time.

Right?

It's happened to Ravelo during our trajectory, I mentioned a couple of times.

So we were sitting on this giant network of hundreds of thousands of pre vetted engineers at a time when the world went remote.

So we essentially, and Ben, I wish I could say I was a visionary and I saw this opportunity coming, et cetera.

Honestly, we're just there, the right place, the right time and we worked very hard to capitalize on it.

Now there are several companies who today are sitting on very valuable assets for this new age of AI.

And those assets are usually quality data, unique data, data that their competitors don't have, proprietary data.

So for me, Roland, so for me, luck in the age of AI is being blessed with the data already being set up and in place.

Because if you already have the data pipeline set up, if you already have the struct your data in place and your competitors don't, you have an immense edge against them for this new age.

So what do leaders need to do?

So what do leaders need to do to get lucky with AI?

I think leaders who got lucky with AI are probably the one who, probably the ones who did their homework ahead of time.

I think that what they need to do now is look at their assets, look across their companies and see what types of data they have that's already prepared and structured and up to date and you know, types of data that their clients or their competitors don't have and that the clients want.

Right?

So that's for me is, is the key, the key issue.

I love that.

So it's like bring an awareness to what you have inside your organization.

If you've got a hundred people, if you've got five people, if you got a hundred thousand people, you are sitting on data.

The question is, do you know where it is?

Do you know what it is?

And then are you actually getting curious about how it can be utilized and implemented?

And I love that.

It's like creating your own luck.

That's it.

And can you link that data to client needs that you can serve better?

Right?

Can you, can you find ways, can you create use cases for that data that will wow your customer?

For me, that's, that's luck.

Okay?

So I don't want to give the Whole playbook away here because I think we just wrote a whole book about luck and AI.

But so, okay, imagine a leader, they're discovering the, you know, this pockets of information they have and some might be on a Tandy 1000 for a Commodore 60, hopefully not a Commodore 64.

They're fighting, they're fighting floppy disk.

They're finding this.

What do they need to be telling their or asking their leaders inside the organization after they find it, like you said, linking it back to customers to value.

Maybe rough on that or maybe like if you have some examples of where this has happened.

I will give an example that happened to us.

Like I'll give the example of Fro Fellow.

When Covid hit in 2020, the data that we were sitting in that was proprietary was the skill sets of the people who were part of our network.

So for somebody to participate in Ravello's network, they had to take assessments and they had to share with us a lot of data about their preferred programming languages, their preferred frameworks, how much experience they had with that, their GitHub profiles, and really things that they wouldn't put on LinkedIn or they would not only get information.

Yes.

No, I mean the recruiters would get to that kind of information maybe at the third or fourth round of interviews.

And we had all that for hundreds of thousands of people.

And you're like, oh, okay, our clients want that.

I can actually save my client time if I can package that information and show them shortlists upfront only with the people that have the relevant skills for what they were trying to solve.

So basically that's us trying to be intelligent about using the information, the data that we had.

And then now that's phase one.

Once you have an AI layer on top of that, you can now say things like, oh, so this person knows this or that framework.

And then there's new tool that barely anyone knows about.

But with AI, I can say this is similar to that framework and that person would be a prime candidate for learning this new skill.

And then I can work on career development for people.

So that's, that's how I would, I would plug these things together.

So if you had not discovered that, would Ravelo be what it is today?

I think we'd be a recruiter serving Latin American companies with tech talent still.

Again, it's a matter of luck, of having the assets.

But there is immense amounts of hard work to making that luck become reality.

So you need to capitalize on the assets you have.

We moved quickly, we made difficult trade offs, we let Go of parts of the business that we were very excited about, to focus on what we thought was going to be the future.

And we were able to be successful doing that.

Well, congratulations on that.

And I like the leaders.

I just think the message underlying this is there's so much hype about AI being external.

I need to go work externally.

But if you go do that and you haven't done the inner work inside your company to understand what those assets are, what your data is, you're going to miss something or you're going to come out kind of like a bland, vanilla opportunity for your customers and not something truly specialized and specific that only you can offer.

Yeah, I think there's.

And you can.

You're correct.

And there's a few waves of that that have happened during our trajectory.

For example, over the past year, the demand for human data, that's how they call the data that is created by humans to feed and post train those LLMs.

The demand for human data has gone through the roof.

Now there are companies who could just provide the humans, and there's companies who have the expertise about generating human data, who have built the platform like we did and who have, you know, tailored their approach to what the hyperscalers want.

And I think that it would have been easy for us just to coast on the fact, hey, we have the huge network of talent.

I think that would be a path to mediocrity.

I think that if we hadn't tried try to push one level deeper and try to understand how can we serve these clients even better, I think we'd be in a bad place.

So I alluded to some of this in the beginning, but you've done what most people dream of.

You've been to elite schools, Stanford, top firms like Goldman and Bain, and you've got a long list of startup wins.

But what part of that actually mattered when things got really hard?

That's a great question.

Part of me wants to say, well, the brand names don't help a lot because when the hard parts of the business, the hard parts of running your own company, they're not very glamorous.

And, you know, you can't cling on to brand names when that happens.

However, going through those experiences, I think teaches you humility because you get in touch with people that are so much smarter than you and that, you know, are so much more talented than you, and you realize that you just need to keep working hard to get where you want to get.

So that humility that going through those places teaches you, that grounds you when you are like having a terrible week at an early stage startup and, you know, running out of money and your client, you have a, you know, client issues or something like that, you remember, look, I, I felt as if I was in, out of my depth.

I was out of my depth when I was going through some of these places.

The feeling's familiar, right?

So that familiarity, the family of the, the, the feeling of not having, being.

The feeling of not going through this for the first time, I think is actually very helpful.

So, so yeah, I think it's a long answer to a short question, but that's it.

I mean, it teaches you resilience and humility.

I like that.

And it, one of these words that comes to mind too is courage.

Because it's not easy.

You know, there's like, there's the adage of, well, if you're, if you're the smartest person in the room, you're in the wrong room and all this other stuff.

But it's hard to be in a room where you are not the smartest.

I'm not saying you weren't the smartest at Stanford.

You might have been.

Oh, so many rooms.

Yeah.

Okay.

So to keep, so to keep putting yourself in that room where you are humbled and to show up and then also to me is to have the courage to engage with people even if you don't.

You know, it's easy to engage with people when you have the answer.

It's a lot more humbling and harder when you don't.

So it sounds like you've been able to do that time and time again.

Show up, engage.

But the flip side, you know, you mentioned, you know, challenging times in your career, running out of money, you know, having a lot of uncertainty, making huge bets for the company.

What, on a personal level, what do you do in those moments?

Like what, what tools?

Process people, tactics.

She just like, yeah, what is it?

How do you deal with.

Just felt like you were describing my, my, my, the last 10 years of my career because it feels like I'm putting myself in that position of, you know, not really having a clue about, about things and you know, not at all close.

Being the smartest person.

Yeah, they're like, that's the space you're in.

You're not in this space.

Everybody has all the answers.

If you want that, you should go different Bill.

That's it.

And I, I think personally what grounds me there is to one, work is not the only pillar of my life.

I think that having strong pillars in your life, family, friends, Spirituality, sports, health.

I think that's, that's grounding and it helps, helps you keep afloat and stable.

And two, I like the mindset of staying a student.

I think that staying a student, being somebody who can continuously learn is something that again grounds me during those times because if you look at a situation where you're in out of your depth but you think you reframe it as okay, this is how I learned to swim.

This is, this is a great way to get through this.

I think it's, it's personally how I, how I go through these things.

Yeah.

Having diversification outside of work don't make work, just your life.

Although in your business it's probably hard not to do that.

You don't very consuming.

What are you, what are some of your go to's outside of work that, that are helpful?

I think my family is the most important one.

I spend a lot of time with my son.

I have a 4 year old son and I spend as much time with him as I can because I know that's going to go away soon.

I mean he, 10 years from now he, he's not going to want to spend any time with me.

I, I swim a lot.

I like being in the water and you know, just like it's my, my way of meditating.

When you're in the water for two, three hours pools or talk to anyone.

Yeah.

But open water as much as possible and then, and then pools.

Yeah.

Do you listen to, do you listen to podcasts or music while you swim or you just.

No, no, just the bubbles.

Yeah, just the bubbles, yeah.

A good friend of mine, I've tried to do long distance swimming but it's just my mind starts playing games on me and I've started to listen to things and all that and I have a friend who just goes, goes swim, swim, swims in pools over and over and over and occasionally does open water.

But he's like, yeah, think about my kids.

I have this numbering system that I do every time and you know, for him it's like a mental sort of gymnastics he does.

But it's a. I think the first 10 or 15 minutes are the most challenging.

After that you just get in the zone and then you, you're in your mind solving problems and rethinking things and you have ideas, you throw those ideas away, you re rehash those ideas.

I think it's really, it's really important.

Kind of part of my routine.

Lucas, this has been a fun one and I just don't think the world's thinking about luck and AI and creating your own luck with AI nearly enough.

But with that in mind, let's.

I'm going to let you have the last word here.

Maybe something you didn't get to talk about that you'd like to mention or you can recap an idea or, you know, tell a story any way you want to go.

I like look, thanks.

I. I like that we are living at a time when no one's certain about the future.

I think they're more than any other time in history.

There's.

There's one, only one certainty and that's things are going to change.

I think what we're doing at Ravello is trying to position ourselves as enablers of that change.

So instead of trying to get where the.

Yes.

Where the future is going to go, when I say that we are building the backbone of tech talent for the age of AI or for the opex of AI, it's just that we think that those are the trends that are going to drive the future.

So if we can connect ourselves to those trends and empower them in some way, I think we're going to be well placed for whatever that future is.

I like that way of thinking and I think that many CEOs could benefit from that.

You don't really know where things are going and if you have to pinpoint where it is to get right, to be successful, then you're probably going to lose.

So how do you plug yourself into the trends that are going to drive that future?

It really changes goal setting for companies.

It's like, yeah, I want to hit this revenue threshold, selling this service or this one thing to these people.

And that target may very well change.

And so this idea of it's like, hey, put yourself in the position to be flexible enough to actually hit the target that matters a year from now.

And we don't exactly know where that's going to be.

That's exactly it.

Lucas, thank you for coming on, my friend.

Good.

Thank you, Ben.

Thanks for having me.