We are opening a new Cloud region here in Mexico in Querétaro. It is a big step in bringing our technology to Mexico. In the country, we are also announcing as part of that an investment plan to train 1 million people in the country so that they can learn the technology and also use it to build new applications and modernize technology for companies and the government. For us, just as an organization, Mexico is not just a market in which we bring our technology to companies, but we also have engineering. For example, our Technical Support Center, our technical account management globals team is based here in Mexico, and it's because of the talent we see in the engineering community, and particularly the younger people who are coming out of universities. What we see a lot of companies wanting to do is using the technical skills of the students graduating from university in computer science, for example, to then build the next generation of technology. We are hoping that we bring two groups of people: some of them will be engineers who want to learn the new thing, AI, but also people who are not necessarily technical because our AI tools are making things much simpler for people, so you don't have to be a real engineer.
Thomas, thank you very much. What are the details of the announcement today in Mexico and what is the importance of the announcement?
We are opening a new Cloud region here in Mexico in Querétaro. It is a big step in bringing our technology to Mexico. In the country, we are also announcing as part of that an investment plan to train 1 million people in the country so that they can learn the technology and also use it to build new applications and modernize technology for companies and the government. Finally, we've done studies that show that between now and 2030, over the next 5 years, this big new Cloud region can expand the economy of Mexico significantly by 226 billion pesos. It's a continued commitment from Google. We've been in Mexico, as you know, for a long, long time—19 years now—and it's a continued commitment to expand our presence here in the country.
So the 226 billion should be the outcome in terms of the economy until 2030. Besides the investments on the cloud region, the infrastructure—what is a region? Because I know what it is, but maybe it's very useful that you please explain what is the difference between a data center and a region.
A region for us consists of multiple data centers. Each data center is called a zone, and so they are then connected together with a high-speed network so that you can build an application and a system that works across multiple data centers. The reason is it gives you better reliability, it gives you better security. So our region is actually not one data center; it's a collection of data centers. And there are 40 regions of Google Cloud around the world. This is the 41st.
Yes, the 41st. And what is the importance in terms of—because talking about 41 regions now, the Mexican one—what is the size in terms of the different sizes of regions, or are all the same?
Almost all our regions have standard configurations, you know, in terms of both the design of them and then the services we offer. Now, because many companies here in Mexico want to sell, like Mercado Libre, for example, as a customer, they are in many different parts of the world, including in different parts of Latin America. They want to build services that work consistently across our regions, so our regions are fairly standard in configuration. In Latin America, Brazil and Mexico are the only regions. Oh, you also have Chile. Chile, that's very interesting for services out of Chile.
What is the challenge in terms of the infrastructure? Because energy is a big issue in this, in terms of having renewables. We have seen many deals in terms of nuclear energy in terms of Mexico. What is the main challenge in terms of energy, or how could you face it?
We have many different pieces that go to make a data center or a cloud region. There is, you know, obviously we need space to put the buildings in. We have to deploy renewable energy and resilient—many times when you look at renewable energy, there are sources like solar and hydroelectric and other things, but because the region needs to operate 24/7, we also have to have other redundant sources of energy. So we work through many of those solutions. We need to deploy very low-latency fiber network, not just to connect the region and the data centers together, but also to connect them to the rest of the world, so to connect to our backbone network, for example. And so these are things we do. In fact, people working on this—we have been under development for a couple of years now, so we are super excited today to bring it to everybody.
Yeah, it's very important for Mexico because also it's infrastructure of the present and the future, right? Because AI, etc. Now, in the Mexican region that you're announcing today, what is the percentage in terms of AI? Because our business in Mexico needs a cloud just for the normal business process, but what about the GenAI, and what is going to be the role in the Mexican region?
AI is being used by many companies that we do business with in Mexico, from banks to insurance companies to retailers, etc. AI, for example, is being used to help retailers provide consumers the ability to find products easily, you know, to search and find products, to improve customer service. For example, we have programs at Liverpool, FEMSA, a number of them. We also are supporting Nubank to build applications in financial services. We have work going on with BBVA, for example. So our work spans this, and many organizations are starting to use AI to improve productivity internally, to streamline their internal processes, to serve customers better. And we already have them using these AI services in our other regions. Now you can use them here.
And the foundation model is going to be in the Mexican region, or it goes to the US probably?
Probably our foundation models—we roll it out typically after a region is launched. Within a few months, we bring the foundation model to the country. Our foundation model is called Gemini, and so we have a lot of customers interested in using it, and you will see it here too.
What are you seeing in Mexico in terms of—usually I was listening to a conference of you and you were saying we're transitioning from IT initiatives in terms of going to the cloud and GenAI now to business initiatives. From the point of view of business, how are you seeing the Mexican landscape in terms of—is it departments or the CEO or someone on the upper management saying we need to go into the cloud in order to capture GenAI and all the solutions?
We see both patterns. In some companies, they start with 'I need to modernize my IT system, so I should go,' typically led by a CIO or CTO. In other companies, we are seeing the head of customer service, the head of e-commerce, the head of financial regulation, for example. There are financial institutions looking at using AI to improve regulatory reporting and fraud detection. For example, there are people using it in the back office to streamline the back office, customer help desk, internal help desk for employees, etc. Those tend to be driven by the line of business. And so we see both patterns going on. In some of the leading companies here in Mexico, we meet with the CEOs, and they are very interested in using AI because they think it's a way to take the company forward and to leapfrog the competition. So we see that too.
Will you say that the GPUs are going to eat all the processing? Because now we're seeing the GenAI is a transversal process, so it's going to be all the processes, I guess, in a couple of years, or how—what is your vision in terms of that?
We continue to make our models extremely efficient. You know, we can run our models even on a small phone now. So, to train a model, you still need accelerator chips, but increasingly we are seeing models becoming smaller and smaller while maintaining super high quality. And so more and more we see the combination of a GPU or TPU, which is our accelerator, plus a traditional compute processor. And we see a lot of applications being built to exploit both.
What would you say is the take that a country like Mexico, which is an emerging country, a big country, one of the 20 biggest economies—what should we see in terms of the mindset as a country in order to take advantage of this new way of technology and AI?
Great question. In Mexico, for us, just as an organization, Mexico is not just a market in which we bring our technology to companies, but we also have engineering. For example, our Technical Support Center, our technical account management globals team is based here in Mexico, and it's because of the talent we see in the engineering community, and particularly the younger people who are coming out of universities. What we see a lot of companies wanting to do is using the technical skills of the students graduating from university in computer science, for example, to then build the next generation of technology. And the benefit you now have is the same AI that sits underneath Google Search or under Google Maps is now available to these customers. So we see companies now starting to reimagine—and when I say reimagine, for example, if you're talking about customer service, how can you allow a customer on WhatsApp to talk to your team but not talking always to a human? Because sometimes in the day they may not be people at the call center, and you can still check, but it's answered by an AI system. And the AI system can be extremely accurate in answering questions, helping customers. So that's an example of things we're seeing. We see companies using it, for example, to help people buy things, you know, assisted shopping. If you want to return something, instead of having to talk to a person, you can have a system answer your questions, you know, help you figure out how to return something if you want to. All of these are examples of things that new companies are jumping on and learning how to do with AI. To do AI well, you need both the technology, which we are bringing, but also the expertise. And the opportunity we see from Mexico is that your engineering student population is learning all of this new technology. That's why we're investing in a million people to get certified. I think it gives the opportunity to leapfrog.
Would you say that—and talking about leapfrog—you need to be an engineer to be part of these million persons that are going to be prepared in terms of the certification, or we should imagine other ways of curriculum and education?
It's a great question. Some of the people—people will want—some of the—we are hoping that we bring two groups of people: some of them will be engineers who want to learn the new thing, AI, but also people who are not necessarily technical because our AI tools are making things much simpler for people, so you don't have to be a real engineer. For example, there are lots of people who are not hardcore engineers using and building, using our AI platform, new systems because AI can simplify technology.
