Manoj SharmaFor two decades, Mexico’s outsourcing story revolved around two things — headcount and dollar volume: how many seats some of the big India Inc. vendors added or how much cheaper a Mexican coder is compared to an American techie. In recent times, this model has become obsolete as Mexico transitions into a new era defined by specialization and adaptability.
Mexico’s IT services industry has reached $21 billion annually, driven by North American nearshoring demand, and will continue to grow aggressively, with projections of hitting $60 billion by 2033.
Enrique Cortés Rello, National Leader Strategic Initiative AI/Director AI Hub, Tecnológico de Monterrey University in Mexico, is someone who has been on both sides of the aisle: he watched the Indian IT model up close in Bangalore during its boom time before he returned to Mexico to spend years inside its Latin American expansion.
As he leads the AI initiatives at one of the top universities in Mexico, he argues the country’s real moat lies in refining the Indian outsourcing model — moving from commodity programming to genuinely complex work. That, he says, is defining Chapter 2 of Mexico’s IT services industry, though it’s not fully there.
In this free-wheeling conversation with Nearshore Americas, Cortés Rello talks about Mexico’s rise as a top nearshoring destination in LatAm, why big tech giants like Google and Microsoft are leveraging it not just as an outsourcing destination but to explore new frontiers, and why outcome-based delivery is breaking the old business model. Edited excerpts.
Manoj: Let’s kick off with your understanding of India’s outsourcing model, your experience with Wipro, and before the Mexico story actually took off. What good things did Mexico pick from the Indian model, and how did it do it differently?
Enrique: I lived in Bangalore, India, for three years while I was working with Wipro Systems, which had a joint venture with HP. I was working for the American partner, Federal Systems, so I saw closely how the Indian model worked.
Later in life, I worked for Wipro, so I saw what we produced in Latin America. And I know a lot of Indian friends who moved back and forth from India to Latin America. So I have a very, I won’t say unique, but different perspective — I saw the Indian model in India, then I saw the Indian model in Latin America from a more detached point of view.
When I lived in India 20 years ago, I saw the classic outsourcing model — based on hiring a number of people, training them, and a few efficient people managing them. When I came back to Mexico, I saw a wave of Indian companies, TCS, Wipro, HCL, and a bunch of other small Indian companies, moving to Mexico and facing challenges.
Mexican people have different motivations, and their attitude towards work is also different. You can’t apply an Indian model straight. For example, in India, people were proud to work for Tata. That same model doesn’t translate to Mexico. People prefer to have some sort of work-life balance.
— Enrique Cortés Rello, Director, AI Hub, Tecnológico de Monterrey University
Mexican people have different motivations, and their attitude towards work is also different. You can’t apply an Indian model straight. For example, in India, people were proud to work for Tata. That same model doesn’t translate to Mexico. People prefer to have some sort of work-life balance.
It took some time for Indian companies to understand different ways of working. When Indian companies landed in Mexico, most of the management was Indian. They realized, “I need a middle manager who is not Indian, so that I can deal with this situation.” And they started forming local Mexican middle management, which helped.
The other thing was the Mexican labor laws, which are different from those in India or America. They are protective of workers. Here, you can’t just fire somebody, so these were the learnings and companies adapted. India was — not now, probably — a labor arbitrage destination earlier, and Indian companies came to Latin America thinking of it as a low labor cost destination. But they realized that Latin America is not as cheap as India — cheaper than America, but it’s a middle ground. So they had to adapt their economic models.
There were a lot of Indians in Mexico earlier — not only in management but also tech workers — and all those eventually transmigrated into the local population. A huge number of people work for TCS in Mexico, big numbers for Wipro and Infosys, and HCL. They adapted well, and Mexicans see them as locals.
Manoj: From your understanding of the industry, if we stop measuring Mexico’s outsourcing success by headcount or revenue, what would you measure it by instead?
Enrique: We (Mexico) started as a commodity player: it’s the lowest level of commodity programming, let’s say. Mexico has come up the chain of complexity: from simpler to more complex and valuable things. TCS, for example, does more complex things in Mexico now than it did earlier. That’s a matter of time, education, and the labor force being more sophisticated. I would measure the success by looking at the type of tasks the Indian companies are doing out of Mexico and how they went up the chain of value.
Manoj: And how do you see the two sides: Indian companies moving into Mexico, and the performance of homegrown Mexican companies? Which of these have performed well?
Enrique: There are some Mexican companies that are like the Indian companies — IT outsourcing programming, but not many. Probably, the biggest name that comes to mind is Softtek. Some Mexican companies were acquired: EPAM, for example, which bought a large Mexican company.
Another phenomenon is that American companies are opening engineering/captive centres in Mexico. Oracle has a huge one in Guadalajara, an engineering center. Google has a big one in Mexico City. Apple doesn’t, but Microsoft also has an engineering center.
One more trend I have observed is that many younger, bright professionals choose not to work for companies but as contractors from Mexico for American companies, and get paid in dollars. Very good for them, but that makes the labor market more expensive.
Manoj: What are the metrics the industry celebrated earlier, back when you used to work in India and then moved to Mexico, but are kind of outdated in current times?
Enrique: Well, a long time ago, we all cared about KLOCs — thousand lines of code. Now, here in Mexico — everybody uses AI tools to produce software, everybody. So the game is changing, productivity is changing, and — how can I put it — we need fewer people to do the same thing.
However, when you use those AI tools, other skills become very important: systems integration, testing, system-type testing, and putting things together. Skill sets are changing everywhere: be it Mexico, the US, or India. Methods of success are also different: it’s more about code generation using AI tools, but making sure the code is not crazy.
Manoj: I’m sure there’s also a shift in how companies now look at headcount. Do you also see an approach towards outcome-based project delivery?
Enrique: That’s right, and that introduces big uncertainty, especially for companies used to large teams, large numbers, and linear scaling: more dollars, one more guy. That’s dying; now we’re outcome-based. But outcome-based breaks the business model of “I hire a lot, I train a lot, then they become productive”.
I don’t think we’re at a place where the old model is dead and the new model is fully active, but we’re in that transition.
If you see the QS Global Ranking for AI as a subject, Tec de Monterrey is in the top 100 worldwide. It is number one in Mexico and even in Latin America. So we should be happy about it, but we’re not because the fields or specialties where we compete are not the ones that need a lot of investment.
— Enrique Cortés Rello, Director, AI Hub, Tecnológico de Monterrey University
Manoj: Tell me what’s happening at Tec de Monterrey’s classrooms and in terms of research — you’re leading one of the biggest AI initiatives at the university. What do you think the industry isn’t paying enough attention to right now?
Enrique: If you see the QS Global Ranking for AI as a subject, Tec de Monterrey is in the top 100 worldwide. It is number one in Mexico and even in Latin America.
So we should be happy about it, but we’re not because the fields or specialties where we compete are not the ones that need a lot of investment. We can’t compete with people who have giant GPU farms. So we chose to compete in a different area of research, which is small language models specialized for industry. We take an industrial pain point, apply intelligence, and measure results in a quantitative way — that’s where we specialize.
We’ve chosen a few industry chains to complete: one is consumer products and retail. Another is the construction of houses from raw materials to end-product houses. We compete in those industries because not many understand them end-to-end.
In the university context, for every specialty — medicine, law, economics — the university is trying to determine what they should know about AI in their area. How will the task of a lawyer change when you introduce AI, and what are the tools you will use in your profession?
Manoj: You’ve been on both sides — inside a global provider and academia. What do you think one side consistently gets wrong about the other?
Enrique: I love the question. The problem isn’t what they don’t understand — it’s that we need to align our motivations. What we want in industry is speed, and the motivations of the university are: “I create new knowledge, I don’t need to go fast, and my reward is writing papers”.
The big challenge for universities is to convince industry that attacking big research problems will make them money. You need to solve some very tough research problems, and the guys who know how to fix them are the universities.
A bunch of startups in Mexico are using new technology, but we don’t have many unicorns — we have fewer than India or any major country. I think that’s a timing issue, and we’ll see more unicorns. We are going up the complexity chain technically.
— Enrique Cortés Rello, Director, AI Hub, Tecnológico de Monterrey University
Manoj: If we’ve seen Chapter 1 of Mexico’s outsourcing success so far, what does Chapter 2 look like for this country?
Enrique: That’s a very good question. A bunch of startups in Mexico are using new technology, but we don’t have many unicorns — we have fewer than India or any major country. I think that’s a timing issue, and we’ll see more unicorns.
We are going up the complexity chain technically: so we will see the Indian companies, and the Eastern Europeans, and the Americans, doing much more complex things out of their Mexican operations.
For example, the pharma company AstraZeneca has an engineering center in Mexico. It started as a usual programming center, but now they are exploring quantum computing for pharma from Mexico.
Oracle is doing complex things out of Mexico, Google is doing similar things, and even IBM is doing very interesting things out of Mexico. We are close to America, so we have access to many things that are in America. So that’s chapter two: complex things out of Mexico.
Add comment