 Hello everyone and welcome back. We are kicking off day two of the Cube's live coverage of Google Cloud Next. I'm your host, Rebecca Knight, along with my co-host, John Furrier. John, this is wall-to-wall coverage, Gen AI all the time. It's a great show. It's an industry event in the sense that the demand for Gen AI infrastructure is so high and the user experiences have changed with the Gen AI wave with ChatGPT and now you've got user experiences and the underlying infrastructure has to deliver the performance. So everyone in the enterprise is actually running really hard to re-tool, reset, rethink, and re-architect the engineering behind how to scale up these platforms to serve the demand and of course the developer community is going crazy. They've got a huge appetite for coding, open source and again that's being impacted by AI. So this next segment really hits really the enterprise core product demand for scalable enterprise infrastructure. So it should be great. I was going to say a perfect segue to our next guest. I would like to welcome Sashima Shigatoshi. He is the Chief Technology Officer at Digital Systems and Service Division at Hitachi. Welcome. Thank you. And Srinivas Shankar. He is the Chief Business Officer and Head of Global Industries at GlobalLogic. Welcome both of you. Thank you for having us. Delighted to be here. Srinivas Shankar, I want to start with you. I want you to tell our viewers a little bit about GlobalLogic and also why you're here on the show. Look, let me start with a little bit about GlobalLogic and I'm sure Sami Srinivasan will cover the Hitachi Group as well. Where GlobalLogic is a fully owned subsidiary of the Hitachi Group. We engineer intelligent products, platforms, and services for our clients in ways that have meaningful and relevant positive impact for society on the planet. And we take a tremendous amount of pride in doing that by solving complex engineering challenges. That's essentially who we are, about 30,000 people globally. Look, the reason why we're here is we're really excited to talk about how GlobalLogic and Hitachi are collaborating to create a enterprise-grade, scalable, generative AI platform that will be used broadly by the Hitachi Group of companies. We'll talk a little bit about the approach, what purpose it serves, and what's the importance and relevance of Google Cloud in that purpose. Great, we want to dig into all of that. But Sashima, why don't you start with a bit of an overview about the Hitachi Group? Okay, this year we are honored to celebrate the 150th anniversary of birth of Namihei Odaira, the esteemed founder of Hitachi. He grew our organization to be a global conglomerate now, powering goods through social innovation to realize the sustainable society. Since our founding over a century ago, Hitachi has been committing to its corporate philosophy to contribute society and has been created various innovations. Through the extensive collaboration with a wide range of customers, Hitachi has been accumulating technical and knowledge advances in IT, OT operational technology and product. This is our uniqueness. So these capabilities serve as a key to address the customer's issues and the societal issues. Through what we call the Lumard solution, that means illuminating data to create barriers. So the GNAI is also an important technology for Lumard and the social innovation. For example, as you know, the working age population is a critical societal issue for most of the advanced customer country is facing or will be facing. So GNAI is expected to improve the productivity of our intellectual labor. So that could be a key driver for our customers' businesses and the societies. And the global logic plays a pivotal role as an accelerator of Hitachi Group, which harness the potential of GNAIs. By combining the mutual strength of Hitachi Group, cultivated in the multiple domains in the mission critical field, and the global logic, which has the capability of implementing the advanced technologies, will promote the safe and secure use of the GNAI for our customers and the societies. That's a great vision. And I love that statement because yesterday, one of the big highlights from the show was that the productivity gains are going to free up time. Not just to do more work, but also to do things on their own. But the societal impact is also a user experience. And as the table stakes raise as higher and abstracts away the complexity, we're going to see more change. So a lot of people are doing more transformational things, not just business transformation, but it's enabling a lot of other things. So I have to ask how you guys are building your platform. What is the approach? Everyone's rethinking this engineering and with the data as the enabler, I mean, it's kind of a disruptive enabler because if you have the data done right, the platform engineering side increases. So what are you guys actually building? Take us through the story there. So the essence is that we're building a unified platform. And when I say unified, it does not mean one size fits all. I mean, so much you must have talked about the diverse businesses and domains that the Hitachi group is in, right? So it's not a one size fits all platform. What it essentially is, is a set of foundational assets and capabilities that enable responsible, reliable, reusable implementation of GNI across the enterprise. I mean, each of these three words have very deep meaning and connotation. It ranges from everything, from creating foundational assets, from data ingestion and curation, model selection, training and customization as well as capabilities around security, governance and trust, which are acutely important for an enterprise. Which is why we've actually coined a term for this. We're calling this the platform of platforms. And we truly believe this is a next generation, first of a kind platform or platforms to drive GNI across a very large enterprise like Hitachi. Explain that real quick. What is a platform on platform? How do you, what does that mean from a definitions standpoint? How would you define platform on platforms? Platforms integrating with each other. Platforms on top of each other. What is that, could you just quickly give a quick overview of what platform is like? Let me try and explain that by the analogy of perhaps a three-layered kick. Which is kind of the architecture that we're following for this platform or platforms approach. At the core foundational level, if you think about infrastructure, we're leveraging Hitachi Ventara's hybrid cloud capabilities around reliability and observability to make sure that the platform is secure and reliable. We've also built capabilities and assets on top of that around security governance and trust. I'd argue that nothing is more important for Hitachi and perhaps for any enterprise than trust. And with the advent of generally AI, you also have the dawn of the whole bunch of threat vectors. Never heard of these terms before. Things like malicious prompts that can compromise intellectual property, data poisoning, potentially model drift that can actually give you results that can compromise the trust of the company. So it's a lot about creating these foundational assets that ensure that people like Samashima Saan and I and many other executives are actually sleeping well at night. Right, so that's at the foundational level. You also have reusable and fit-for-purpose language models that are powered by Google's Vertex AI. And also at the top layer, the right set of visualization and consumption capabilities that allow easy adoption of generative AI across the diverse set of companies. So that's really the core of what this platform of platforms approaches. And hence the engineering focus of what you guys do, because that's not easy to pull off. I mean, it's not trivial. Look, it's easier said than done. That's all I'll say. And I think, but we are breaking your ground. And I'm sure Samashima Saan will talk more about the value that it brings to the Hitachi Group of Companies. But I actually want to ask you, Samashima Saan, before we get into that, to talk a little bit about what Serene was talking about earlier. Trust is paramount here. And how are you addressing these governance challenges? Because security, governance, trust, these are critical components. Yes. So as Hitachi being responsible for the critical public infrastructure, we recognize the utmost significance of governance to guarantee the safe and secure use of the AI in mission-critical fields. In order to reinforce our commitment, we established a Generative AI Center last May and gathering the data scientists, AI researchers, which has expertise on the AI, and as well as the experts from IT, in-house IT, and security, quality assurance, intellectual property, and the legal department. So based on our expertise and cultivated in the privacy and the AI ethics guidelines, and as well as the global legal and regulatory trends, we have crafted a comprehensive set of guidelines for the use of Generative AI in Hitachi. And we are now developing the in-house environment to use Generative AI. The entire Hitachi group is embracing the potential of Generative AI by using this guideline and the environment. And so it is necessary to accumulate the practical knowledge when implementing Generative AI. So that's come down to the one more than unified approach as Srinni said. So Srinni, I want to talk about the trust piece coming into governance. She mentioned that the trust at the center, which sounds like the think tank. Everyone gets together to talk about safety, guardrails, all that good stuff, and then how to apply it. But when we look at the scalability of having a unified platform, data has to be addressable. So the big hot area right now, that means it's not really reported much in the press, but governance is hot. People are rethinking governance on how you build that in from the front end. Because if data's going to be flying around everywhere and being generated, you got to know the governance. Can you just share some insights on how you're thinking about the governance piece? Because a three-layer cake to make that work, you got to have a robust governance model. What's the governance strategy? What do you guys do differently? What's the new best practice? Is there a pro tip on governance that you've seen? Look, I think, you know, I'm sure some of you will speak more about how we're instituting governance within the Hitachi Group, but it's really about ensuring that, you know, that the core of it, it's a lot about data. So when you look at data ingestion, you've got to look at the entire data supply chain and ensure there's safety and security, and you're able to govern not just the data that's within the four walls of enterprise, but also the entire, you know, supply chain of data. So that's really critical. We obviously are, you know, creating a library of fit for purpose models, large language models that serve various use cases from conversational AI to field service to knowledge, sharing and retrieval to content generation. And it's really important that you have the right governance mechanisms to determine things like accuracy, especially given that these models are non-deterministic. So what is the right level of accuracy to make sure that you can put your trust of the company behind it? To make sure that you don't have model drift. To make sure that it's grounded in accuracy and it's grounded in relevance and it's grounded in currency. So there's just a lot of foundational elements that you have to put in place to address a lot of that. And I think from an organization standpoint, you know, Samashiv Masan can cover some of the aspects of how do you make sure organizationally we are set up to do that well. By the way, before he gets on there, yesterday Thomas Curry and said in his keynote, the grounding with Google search, that's the Google vertex side and Gemini side. But he said, ground in the enterprise truth. So the grounding with the data becomes a huge deal on that piece there. So that is awesome, awesome piece there, yeah. So as Surini mentioned, the governance of data is a very important point to protect the data or to assure the correctness of the data or something. However, another approach is kind of the architectural or structural approach. One of the hint is in the public infrastructure system. In the control system of public infrastructure, there is a concept named protection control that protects equipment or system independent of the computer system of something. In another word, the guardrail, the word guardrail is used in the AI context. So this kind of the structural scheme is also important, I think. Can you talk a little bit about where Google Cloud comes into this collaboration between GlobalClogic and Hitachi? Well, I believe that Google Cloud is a fast-growing cloud and we Hitachi as a solution providers, we expect Google Cloud to provide the cloud platform and especially the gen AI, general purpose LLM models like Gemini or something. Surini, do you have anything? No, I think, look, to answer that question, I have to say, where do I begin? We're obviously leveraging, as I said, fit for purpose platforms and capabilities that Google Cloud offers like Vertex AI and Gemini Pro. We are actually starting to also understand which models serve what needs better, so we have a real strong point of view around what's good for legacy modernization, what's good for conversational AI. So yeah, we're really looking to leverage the best that the foundational stack that Google Cloud can offer to actually power up the full speed of that three-layered cake, if you will. Well, I mean, yesterday they talked about the first party, third party open source models, which are all now 130 in Vertex, but also they had built on the model guard and they got the model builders and they got the agent builders, so you're starting to see agents come in. So a lot more exciting things are coming. I guess my question would be for you guys is that, what have you learned in your project as you guys are engineering this platform of platforms? What's the key takeaways that you guys could share for folks watching that are thinking, hey, I want the bridge to the future, but I got to build it, I got to engineer it. What learnings are coming out of your work? I mean, you guys were doing this before ChatGPP was launched. So as you pre-launch and now as you're in it, what is the key learnings? Could you share anecdotal things or anything that's not proprietary learnings? Just the key takeaways, right? So you're right in saying that we were, I mean, this is not new for us. We've actually, for many years, we've been serving our clients in terms of training their LLMs. That's actually a service offering that we offer. We actually train LLMs on data in specific domains for our clients. So we've got experience prior to the advent of ChatGBD. Look, I think the key takeaway for me is, 2023 was the year of experimental proof points. 2024 is the year where the industry is going to transition and transcend from experimental proof points to enterprise grade scalable AI that's fit for the enterprise. And I think I'm delighted and proud that GlobalLogic and Hitachi are leading the way on that journey. We're breaking new ground. We're going to be learning new things. You know, the pace of this is so fast that every two weeks, I feel that I'm outdated. New papers are hitting the market. Reading papers on the weekend. Another paper, you got to read. But I will say that we're out there breaking new ground and stay tuned for more. It's great work. And I think it's right in line where the market is now because it is an engineering thing now. It's becoming back to less IT transformation. It's really software engineering transformation. And we're seeing that impact the data levels. I like the platform and platforms because this is what we've been talking about. This is a whole nother level. And it's super exciting. Is there anything you'd like to share to the folks watching around? Advice you'd give them as they go on their journeys as they think about their platform to build. Folks who are out there watching now saying, hey, I want to do this too. What advice would you give them? Okay, let me start. I would say that you have to look at the foundational stack because for any enterprise, you don't want to reinvent the wheel. It's great to say that let every flower bloom. And so you don't want to kind of centralize things that actually limit innovation. But you have to do that to democratize AI. You have to create a set of foundational capabilities that truly then allow every flower to bloom. So that's what we're looking at doing. We're not going to be very prescriptive about how they use our platform. But it's like serving them a set of foundational capabilities that allow them to be secure, that allow trust for the data, that allow them the choice of multiple models in a fit for purpose way. You talked about the agents. We give them a pallet of agents that allow them flexibility around consumption models. So that's what, that's the whole thing. Let them build their own house. Absolutely, and it's about not making sure that we don't reinvent the wheel. Right, no, those are so important. So now that we've discussed how you actually build the technology, how you're architecting it, what are you using it for? And as a future of work journalists, I'm interested on the impact on your business, but also on your workforce too, and how your team is interacting with it. Well, I think the possibility are extensible as the progress of technology is our initial focus is on elevating the productivity by automating tasks, and enhancing the efficiency by fine-tuning the process and growing the revenue, by unlocking the entirely new capabilities. A recent study of consulting firms says that the total economic potential is more than 2.6 or 4.4 trillion dollars, more than the GDP of United Kingdom. So this is, however, merely the dawn of the new era. So we aim to unleash the potential of GNI to achieve the social innovation. One of the things that was really striking when we started talking about Hitachi was this real overarching mission to help society and to create innovations that help humankind. How much does that infiltrate your workforce in the sense of the mission that they're working toward and using GNI to solve some of these big societal problems as well as you were talking about the demographic cliff that we have with the workforce itself? Yes, societal issue is one of the important areas that we need to address and something. The GNI has a strong potential to solve such kind of issue. However, we do not rely on too much on technology itself. We humankind should improve ourselves so that we can determine which is a better answer or which is a better way to go by ourselves. So this is a core evolution of human and technology, I think. Well, both of you seem so positive and bullish about this collaboration. I want you to look into your crystal balls and just talk a little bit about what we might be talking about at next year's Google Cloud Next, what you're most excited about in the coming year. So Rini, do you want to start with you? Look, nobody has a real crystal ball like in hell. I mean, just think about it, two years back. You know, if we were having this conversation, the audience would be like, what are these guys talking about? What are they talking about? I mean, they wouldn't even understand the lingo here that we're talking about, right? So it's really hard to say what's the shape of things to come, but I will say this, that look, if you go back to prior revolutions, you had the industrial revolution. It actually created a 10x, if not 100x, jump in per capita GDP, because it essentially drove a significant acceleration on horsepower. Literally, basically GDP and productivity is a function of how many horses you had. I mean, we have not seen anything like this before. I mean, we're seeing an acceleration of intelligence, right? And it's going to be another 10x multiplier in terms of GDP per capita, productivity, if you will. And we're not even talking about the age of AGI, which is artificial gender intelligence. So I think it'll be supremely empowering. It'll absolutely create new categories. It'll allow us to free our, you know, time to do less mundane work and really do a lot more creative work. And look, I'm really excited about the human in the loop, because the human in the loop in the future is going to be a lot more intelligent and empowered than the human in the loop today. Creativity is going to be off the charts. Yes, yes. And also just the time to live in the world and be with our families. You got a great, you got a great future of work. I mean, it's changing. I do indeed, I do indeed. Sashima Srini, thank you both so much for coming on theCUBE, a great, great conversation. Thank you. Thank you. We enjoyed it. Thank you very much. I'm Rebecca Knight for John Furrier. Stay tuned for more of theCUBE's live coverage of Google Cloud Next. You are watching theCUBE, the leading source for enterprise news.