Red Hat Summit + AnsibleFest 2025 | Maria Bracho, Red Hat

Clips
More from Red Hat Summit + AnsibleFest 2025

Maria Bracho

Chief Technology Officer, LATAM

Red Hat

play_circle_outline Exploring and implementing AI solutions in public sector in Latin America

play_circle_outline Overcoming Challenges in IT Infrastructure and Skills Gap to Achieve AI Success with Quality Data and Training

play_circle_outline Helping companies bridge gap from experimentation to production with Red Hat platform

play_circle_outline Trends in AI adoption, productivity, data utilization, and potential for new business lines

Info
Transcript

Maria Bracho, Red Hat

Maria Bracho

Chief Technology Officer, LATAM Red Hat

Maria Bracho, Chief Technology Officer for Latin America at Red Hat, joins theCUBE’s Rob Strechay and Rebecca Knight at Red Hat Summit 2025 to explore the expanding role of AI in the public sector. The discussion centers on Red Hat’s commitment to ethical, scalable AI, with a special focus on the regulatory dynamics in Latin America.

Bracho delves into explainability in AI and the importance of managing its full life cycle. She examines how open-source models drive both transparency and innovation while addressing the regional skills gap through Red ... Read more

explore Keep Exploring

What considerations are being taken into account when implementing AI technology in the public sector? add

What are some challenges related to implementing AI technology, particularly in the region of LATAM? add

What does Red Hat bring to the table from an AI perspective for companies? add

What are some current trends in the use of data and artificial intelligence within business operations? add

bolt Powered by CUBE AI

Maria Bracho, Red Hat

search

>> Hello everyone, and welcome to theCUBE where we are kicking off three days of wall-to-wall coverage here at Red Hat Summit AnsibleFest 2025. I'm your host Rebecca Knight, alongside my co-host and analyst, Rob Strechay. Rob, it is always so fun to do Red Hat with you, but especially today because we are in our hometown of Boston, Massachusetts.

>> Yeah, I'm so glad they brought it back. I mean, don't get me wrong, we had a lot of fun in Denver last year, but I'm happy to be back here in Boston. I got to do the preview with Shesh a couple weeks ago, so I'm amped up for where we're going this week and what we got to talk about.

>> There's roughly 7,000 people that are going to be descending onto the show floor very soon. A lot of energy in this room, very exciting.

>> And the 10th year theCUBE is here as well.

>> Ah, a milestone here.

>> I know, exactly.

>> So this is perfect way to segue our first guest, a first-time CUBE guest, Maria Bracho. She is the chief technology officer for Latin America at Red Hat. Thank you so much for coming on theCUBE, Maria.

>> Thank you so much, Rebecca. Thank you so much, Rob. Excited to be here, day one at Red Hat Summit in Boston. Great weather, tons of people, a lot of excitement, can't wait.

>> Excellent. Okay, so let's get into this because one of the things that's really great about Red Hat is there's always this real emphasis on practical, scalable solutions that solve real problems and not just innovation for innovation's sake, and that is especially true with AI. I would love to have you talk a little bit about how your customers are exploring and experimenting with AI, especially in the public sector to solve real problems.

>> Yes. I wouldn't say exploration is done. There's definitely a lot of exploration going, but we're moving into practical examples. And in the public sector, they want to know how can they really impact people's life, how can we use this technology for the better, for improvement? And with that comes a lot of responsibility. We're handling privileged information. We want to make sure the keyword has been explainable in Latin America. So there's a lot of focus on making sure that is ethical, that we're handling the data correctly, and that we know what's coming out. And I think that one of the biggest things is just realizing that it's not an input-output scenario, but these AI use cases and examples in the public sector, they have a life cycle of their own. It's not like a traditional application where you build it once and just deploy it anyway and forget about it. But models and data, all of that has a life cycle that keeps changing all the time. So they need to explain all the results.

>> And I think this is something that we talk about quite a bit, and I think it's there's a lot of fun going on when you look at even outside the public sector, but even within the public sector and some of the regulations that are coming down and a lot around data sovereignty, transparency, really looking at how open source is really driving a lot of that transparency as well. What are you seeing in Latin America?

>> Yes, absolutely. So multiple countries in Latin America have started to push out regulations, I would say following quite a lot coming up from Europe, but they're definitely learning from other areas and taking and grabbing the best of everywhere to build them. But in general, there's a lot of consensus of we're handling really privileged data. We really need to do this right, and we need to come up with mechanisms for which we can use and explain this. One of which is leveraging open source, leveraging communities, leveraging examples that others are doing, and the public too is expecting that. So the public at the same time is evaluating AI in a completely different way. So it may mean something for them, and they're going to have higher expectations as they go along.

>> One of the things that you were talking about earlier is explainable AI. I wonder if you could, first of all, explain explainable AI and talk about why it is especially important in Latin America and how it's helping to drive widespread adoption.

>> I think it's explainable everywhere that you have a responsibility to the public to show. That it's a term, explainable AI, and it falls into that. If we see AI models as a black box, of stuff comes in, stuff comes out, and if we start relying on AI or start relying on agents in the same way that we rely on a Google search to give us answers, we're really missing the point. We want to be able to know exactly what's happening in the technology. We really need to understand and make sure that it's transparent how the entire stack is working so you can essentially explain how things are happening.

>> I think one of the things that we've been talking about in the past couple months, all the way back to KubeCon over in London, a lot of things have happened where people are, there's challenges and IT challenges and there are infrastructure challenges getting ready for AI and really helping to modernize their applications because it's not all just about, hey, the best, fastest model, the biggest model. Now we're seeing a lot of smaller models and things more around inference and things. What are you seeing as some of the common challenges from an IT infrastructure perspective that are common across Latin America and others?

>> You would see, well, I think it starts with data and understanding your data and good quality data, AI-ready data, data that you can explain where it came from. So it starts from there. And then you have a huge choice of models. So many models, big and small, open-source models, not so open-source models, understanding the data in which those models were trained to produce certain outcomes. That's huge. And then the other challenge not just related specifically to the technology that it's touching, is the skills gap. We are relying a lot on AI, but understanding that this is all new. I think specifically for LATAM, it provides a moment for this region to leapfrog, to take on and learn new things and then be again, ahead of the curve or at least catching to the curve when it comes to understanding and having more AI skills to apply all these new technologies that we're shipping into results that are practical for people and change lives and move the needle and grow the technology and move forward.

>> So the skills gap, there is a real concern that that could slow adoption. How are you seeing organizations respond?

>> Well, the funny thing is that they're relying on AI to help with the AI skills gap, and there's so many trainings and information and there's the plethora of information everywhere to train and to grow. And Red Hat, we are one of them also providing that kind of help alongside services, alongside certifications. So similar to the same way that we've been doing for Linux and OpenShift and OpenStack and all the products that we're doing.

>> So you hit on something very interesting and it ties into the skills gap a little bit, which is the industries that you see in LATAM that are really maybe further along, like I was talking to a major telco out of Argentina two weeks ago, and they leapfrogged some other even US-based in their use and their adoption of AI and how they were using it. They were using it for fraud detection and some other stuff to get the spam calls down and other things, really neat stuff. What we've found and what we've seen is that in the US some of the industries, like highly regulated industries, have really leapfrogged other industries because they didn't go all cloud-first, and so their data was organized in a... Are you seeing similar trends in LATAM or how-

>> Well, they didn't go all cloud-first also because of the importance of the business that they were talking about. It impacted customers. And they also faced regulation within the industry, but not wide regulation. And then as long as you could pick up your phone and have a phone call, there wasn't really something the public needed to know in terms of how AI was used for Telco. You didn't need to know that that's exactly how they planned where they're going to put out radio access network. But when it comes to GenAI, it's touching people. So some of these answers need to be really crisp, really understand what's happening, and if it involves customer data, you need to be able to take an additional level of care for that. So that's absolutely telco's an industry that's been using AI for a long, long time. But in particular GenAI, they're also having some challenges with skills and trying to adopt it and figuring out the right use cases and figuring out exactly which data and which format the data is going to be consumed to get actionable, practical results.

>> And there's a lot of research that shows that actually having constraints, having regulations put upon you will increase your creativity because you have these boundaries that you need to work within.

>> Exactly. It's like you go to the supermarket to buy eggs. If you have thousands of choices, it'll almost be better to have two or three, at least it gets you going. I think we can be sometimes strapped for choice and basically paralysis by analysis, trying to figure out which one is the best model? Should I do it this way? Should I do it this other way? And the answer is going to be it's the right one for right now because this thing continues to evolve and new models continue to pop up. And what I like about it is that the open-source community is really doing a tremendous amount of work. But to me, the OG open-source community, which is the academic community, is also helping out with new algorithms or ways of processing or doing math and getting us closer to be more effective and better use the hardware. So it's all evolving.

>> You started this conversation by saying exploration isn't dead and that a lot of companies are still doing that, but there is a big leap between going from experimenting to production. How are you at Red Hat helping companies bridge that gap?

>> Absolutely. Well, we really go to companies and we help when they try to ask us like, well, what is Red Hat bringing into the table from an AI perspective? I try to say that it's not too far to what we have brought up before. We are the platform that gives you the ability to have choice of what model you want to use, when you want to use it, what you want to use it for, for what purpose. And the beauty of it is that we've built a platform for containers and virtualization and run applications that these applications are the ones that are being fed by that AI. And you can build that in a platform that looked very, very similar. And we already understand what I think is the next challenge beyond exploration and production, which is the life cycle of these beasts, the model or the same way that we know how to handle life cycle of applications.

>> So I think you hit on a very interesting point there because Red Hat has such a broad portfolio. Where are you helping organizations get started? Because like you were talking about data. One of my favorite open-source things that both Red Hat and others support is InstructLab, where you can go and do synthetic data for doing and building AI. Where do you see people really getting started with Red Hat and with AI?

>> By the way, we see multiple things. We see customers that have already done some exploration on their own with maybe public cloud and larger models. And then they say, "Well, but how do I take this to production? How do I use my own data? And then sometimes my own data isn't enough to move a large language model, so then I need certain tips and tricks and tools to see how good do I really get the output that I want," and that goes back to that black box and how do we get to it? And then within InstructLab, that particular feature of synthetic data generation, you basically inject your own data and it generates additional data that is similar to the one that you inject with so that then you can move a model. And that works especially well with smaller models that can be moved faster with smaller data, but also with large models. So to me, that involves synthetic data generation it's been such a game changer and super proud to be part of, that Red Hat is part of that.

>> So Maria, we're just getting started here at the Red Hat Summit. There's lots of exciting announcements that are coming down the pike, but I want to ask you a bigger, overarching question.

>> Don't ask me about the announcement, please.

>> Oh. No, look-

>> .

>> , right. But I want to ask about the trends that you're seeing in terms of the future and how they will shape how organizations adopt and scale and govern AI over the coming years.

>> So I think the power of community and sharing stories, and this is what Red Hat Summit is all about, we're going to hear about so many customers talking about all their experiences from industries to public sector and satellite data analysis to healthcare to telco, to SSI, of course. So you're going to hear all of these examples, and it's funny that the common thread is the platform that they're using. So I'm excited to hear about all of that. In terms of trends, I think everybody started with a chatbot. We're still going to see some of those, but how to make that smarter with my data and how I can get more insights. And then two trends. One is productivity. How do I help my teams be more productive? So have AI infused in the tools that they use to make them more effective at the tools that they use, and I will stop there before I announce anything. And then the other area is, well, what do we do with all this data that we have? And one trend or one beginning trend that I've started to see is, for example, SSI or specific companies looking at the data that they already have and the business and the market share that they already run into and saying, "Hey, with this data, if I use it this other way, I can absolutely go into new lines of business. I can stop being this company that does A, B, C and just go down the alphabet to many more things that I can use and continue to serve the same customers in the same age." So that's exciting.

>> The world is your oyster.

>> I was going to say, and just some of that stuff, it's really just people get overwhelmed. What are some simple things or a simple place that if somebody wants to get started, where should they go? They're not here this week. They missed out. This place was oversold from what I heard as well. So I mean they were looking for hotel rooms all over the city. So where should people who maybe aren't here this week but want to get started with AI with Red Hat, where should they go and get started?

>> Yes, I think we've made it simple to just encompass everything we're doing with AI with a Red Hat AI basically name. So if you search up that, if that's the one thing that you take, that's going to take you down a rabbit hole and you're going to have a lot of fun finding out. And other than that, I mean there's a ton of content out there, YouTube and others, and it continues to grow. So it's evergreen, ever-growing, and I'm excited to see where we go.

>> Excellent. Maria Bracho, thank you so much for coming on theCUBE.

>> Thank you, Rebecca. Thank you, Rob.

>> Thank you.

>> I'm Rebecca Knight, for Rob Stretchay, stay tuned for more of theCUBE's live coverage of Red Hat Summit AnsibleFest 2025. You're watching theCUBE, the leader in enterprise tech news and analysis.