 Welcome back everyone, theCUBE's live coverage here in Google Next in San Francisco, Moscone. theCUBE here with John Furrier, me and Dustin Kirkland here, co-host of theCUBE. We also have theCUBE team coverage, Rob Scherche, Lisa Martin here. Our next guest is Surab Mishra Global, head of Google, business for Quantify, a breakout company, multiple awards here at the show. Obviously Google Cloud really expanding their ecosystem and you're starting to see a lot of winning companies popping out on the more cloud goodness. Surab, thanks for coming on theCUBE for your time. Thank you so much for inviting me. So before I get into your company and what you do there, I'll show you your role. I saw you at the partner of the year, I had to weave that in. You guys won multiple awards in multiple categories. Quickly give the news on the awards. On the awards, this year was phenomenal for us. We got four awards. The biggest one was around winning, breakthrough, partner of the year for North America. This was our second year in the row. We won the same award last year as well, so we're very excited about it. In addition to that one, we also got award for the expertise that we're building in the manufacturing domain. We did a lot of good work with our customers in the manufacturing industry last year. So we got that award. The next one was around marketing analytics. How can customers of Google Cloud understand their customers better? So that was the other award. And the last one was around talent development. So we know Google, there is a lot of demand for Google, but there are not a lot of practitioners. So we've been able to upscale our team, and we have at this point of time more than one certification per person who focuses on Google Cloud. So congratulations on the breakout thing. Obviously we've been saying on theCUBE, you know, get that data get free, let it be scalable. And really the industry verticals are all impacted. It's very clear, and there's no really debate. Every single industry is impacted by AI. Take a minute to explain what your company does and what your role is at the company. Sure, I can start by first my role, and then I can get into a little bit more about the company as well. Like I said, my name is Saurabh. I lead Quantify's Google Cloud business globally. That's focused on our Google Cloud partnership. When I joined the company back in 2016, one of the co-founders, Asif Hussan, and I started this partnership. Now the team of two people have grown to be a team of 2,200 people, spread across US, Canada, UK, India, and we recently expanded to Singapore as well. Now talking a little bit about the company, we're an award-winning AI-first digital engineering company. We've been premier partner with Google Cloud for seven years now. So AI-first digital engineering company, that's an interesting string of words. What does that mean to you? And out of that fit into the whole ecosystem we have around us here at this Google Met. Sure, definitely. I'm happy to talk a little bit about that. So part of the reason that you've not heard about this, because we believe we are defining this category, where we are helping customers harness the power of data and AI on cloud. Machines today have ability to see things, hear things, understand language, and process patterns. Now using these powers of machine, we are solving for four classes of business problems for our customers. The first one is around Solvage for Knowledge and Discovery. These are systems of search and retrieval built on large corpuses of data. Think of these systems like GPD and Bard, which we are all familiar with. So second class of problems that you're solving for is solving for experiences. Think about speech-based interfaces, gesture-based interfaces, conversational agents. We've all used virtual agents. We've done spoken to voice bots, chat bots. So these are the types of solutions we're building for our customers. The third one is around solving for automation, where we are helping our customers process documents using Google's Document AI, find anomalies in their manufacturing lines using visuals quality inspection. The fourth one being solving for simulation, where we are helping technology to generate synthetic data, build digital twins, industrial metaverse, et cetera. Now these models to build these types of applications are neural nets, which are very data hungry and compute hungry. Now for customers to be able to take advantage of these applied AI solutions that I just talked about, they need to modernize their data fabric and revamp their infrastructure to have GPUs and GPUs to be able to run these models at scale. Now this is our second side of the business, where we are helping customers move their cloud infrastructure from on-prem to cloud. Sorry, move their infrastructure from on-prem to cloud. So that's one, build security around it so these models are secure and are not having any attacks. The second one is around data modernization, where we are helping customers move their data from legacy systems like NetEza, Teradata to Google Cloud, build intelligent data lakes in a way that they can basically take advantage of building these models. The third one is around application modernization, where we are helping customers take their legacy application and rewire them in a cloud native way so they can support the performance that the customers expect. And the last one is around BI modernization, where we are helping customers build dashboards using Looker. I want to get into the role Google Cloud plays in the partnership, obviously, many years, but I want to first, on that first segment you went through there, I wanted to ask, what's the drivers behind that transformation of your customers? Is it because they have an old legacy environment? You heard that in the keynote today, or is it because their data's are all over the place or all of the above? What's the core? We got a lot of services. So what's the driver? Just modernization, what's the main pain points is there any pattern to the problem, or is it more pretty wide? What's the driver? So I would say there are definitely a few themes, right? Over the years, a lot of our customers have built different systems, have acquired different data sources that are basically in silos, in disparate sections, right? Now for us, for them to be able to build a robust system, they need to be able to bring their data in a way that those data pieces can speak to each other, right? So we have a unified data record for us to be able to build intelligent systems on top of that. This is one of the very common themes that we are seeing with the customers that are starting on their digital transformation journey. So that's the first part of it. The second part of it is the customers who are already on cloud, who are already currently on the legacy systems, right? Those customers, for them to be able to harness the power of AI, they're starting to move to cloud, right? And then if your models are sitting on cloud, they would prefer to eventually move their data to cloud as well. So we're seeing a lot of these customers move their data from legacy systems to Google Cloud, BigQuery, et cetera. So those are some of the patterns that we're seeing. Yeah, that's a great lead into the next question, which is around the Google Cloud collaboration. Where do you utilize Google? How do your customers experience that? And what problems does Google Cloud solve for you? That's actually a very good question. Google Cloud is actually a core to our business, and it helps us in multitude of ways, right? So first of it, Google helps us provide scalable and reliable infrastructure so we can build solutions, the AI systems that I talked about, that are high-performance and are basically giving customers what they need, right? So that's the first, the foremost part of it. They're helping us with the infrastructure side of things. The second part being around data and AI services that they offer for partners like us to build these systems. Two of them that I mentioned is Vertex AI and BigQuery. Now, these services provide us technologies that help us build these sophisticated models, understand the large corpus of data, and make meaning out of it. So that's the second part where Google is helping us with these technologies, right? The third part is a large variety of pre-trained models, APIs, and developmental assets, which are able to help us accelerate our journey to Cloud. So this is how I believe Google fits into it. They've been instrumental to us. Great, great. Pardon John, Google's obviously positioned well. Dustin used to work there. He's an ex-Googler. He knows. A lot of AI jobs, obviously bringing that to the table, which we heard a lot of the executives that we've interviewed, they're bringing up the best of Google's jewels into the clouds, that's cool, they're cloudifying it. But the big upside is the generative AI conversation. So can you talk about Quantify's view and role in contributing to this wave? Because developers love it, the boardroom loves it, the minds of business loves it, software loves AI. So what are the opportunities and challenges that you guys are pursuing with generative AI? Sure, definitely. So let me, so state of AI has evolved significantly over the decades, right? Now I'll go back to the days of pre-TensorFlow, which was open sourced by Google in 2015, right? AI was limited to mostly structured data in rows and columns and predominantly used for doing predictive analytics. With AI, with TensorFlow, the new class of AI came across, which is around deep learning, which gave machines the ability to see things, hear things, understand language, right? Now it's been five to seven years that TensorFlow was launched, but these models were still cumbersome and expensive, right? Because you had to develop a separate model for every cognitive task. With task, right? Now building these models was cumbersome. Now as the AI ecosystem evolved, with valuable contributions from open source, the art of the possible will AI expand, right? I still remember getting our hands on Google's research paper around generative network architecture. That was one of the big things that happened for the AI community and we knew it back then. It was obvious, the paper. Yeah, exactly, we knew it back then, right? That this is going to be revolutionary and it's only a matter of time that the impact of generative AI will become real for customers. And because you also mentioned like how does Google helped in this journey? I would say that TensorFlow was open sourced by Google. This research was published by Google and these two things have basically formed basis of how generative AI has happened over the years. Yeah, so the positive aspects, the possibilities are endless. What are some of the limitations and challenges of this technology from your seat and where you see it? Thanks for bringing that question, right? So these generative AI, everybody's talking today in today's world about the hype of the possibility of what generative AI can do. But there are a lot of challenges and people who are leveraging this technology need to understand that and basically address those things and use this technology in a little bit of a responsible way. Now talking about some of the challenges that we are seeing the first one is around accuracy. Now these models are trained on large copies of data and follow complex patterns. So they have a tendency to give erroneous results or what we call hallucinations, right? So these answers may be good semantically but when put into a context in which the question is being asked, they may actually not be correct and which is basically questioning the reliability of these generative AI. I think all of us at this point have asked the AI an obvious question and gotten a straight face lie to our face at this point and it breaks a little bit of the trust that we have in it. So what are you doing about that? So for this, what we are doing is we are basically, we've been working on ethical principles of responsible AI. So the systems that we are building, we are ensuring they are free of bias. They are fair, they are transparent and explainable, and they're secure and have privacy baked into it. So these generative AI models that we are developing for our customers are being used by them in a way that they are intended to do so. So these are some of the principles that we follow. They are like eight principles that we have curated in our labs and that's what we are basically doing. We've actually published a paper on that one as well. There's a huge skill gaps too and depending on which age group you talk about also the younger generation are all over this. I mean the startups in Silicon Valley alone is like highly concentrated. There's more meetups than ever before. There's definitely excitement in the open source. So you're seeing a lot of action. What's percolating up in your view that you see from as AI starts to hit that first organic wave of adoption? You're starting to see like Lang chain out there and you got Google's got embeddings. I mean integrations. It's kind of like an interesting formation going on. Will something have to emerge out of this wave before the big enterprises jump on it or people's going to ride their data on their way or how do you guys see that adoption? Because we're seeing a lot of experimentation. Not a lot of production mostly integrated into an app but not a lot of net new production yet. So we expect it to be coming. So for this question, actually this is a good question because again going back to the hype that the generative AI has created we're having a lot of conversations. So I will tell you in the last four months we've spoken to 300 plus customers who want to understand the possibility of generative AI. So what we are doing to help customers embark on this journey is first we build like a four step process for them. The first one is help them demystify this generative AI. What does it mean for them? And what does it mean for their industry? So that's the first part of it. Now with our conversations that we've had with so many customers we've been able to come up with some patterns of what customers are trying to solve for. The second one is basically doing the rapid prototyping the thing that you mentioned about experimentation. Now what we do is we work with these customers to first curate the use cases that can be solved for their industries because we want to make sure that they have the right data set to be able to solve the problem that they're looking to solve for. They have the right infrastructure to be able to do so. Now once the prototype is successful there comes the big part which is what your main question was, right? How will you see this experimentation go into adoption? Now for this to happen there have to be some considerations where we have a package for our customers where how do you take the prototype to production? We talk about the data considerations we talk about the infrastructure considerations we talk about the security considerations which is very, very important. Then the other two which have become very, very important in context of generative AI is legal implications. And the last one is around change management which people just normally end up missing about because these systems are going to fundamentally change how industries function. And all the process. There is going to be a lot of effort that needs to go on change management for these systems to actually get adopted within the enterprises. So very briefly can you tell us specifically what are your product offerings in the generative AI space? Sure, so today our generative AI offerings revolve around integrating generative AI for domain specific tasks to enhance productivity for knowledge workers. At the forefront is our enterprise generative AI platform which we are calling as Bionic which is a combination of Bionic plus AI plus Quantify and Bionic enables enterprise to access and leverage fundamental models that are available in the market to be adapted for their own domain to solve wide variety of use cases. What's unique about this platform is that this enables customers to plug in their own data, instruction fine tune them for their own tasks that they are looking for to bring generative AI capabilities within this organization. In the true sense, what we believe the Bionic is helping customers unleash the power of LLMs and do this in a safer way within their environment. So that's basically our offering Bionic and we're helping customers get onboarded on this journey. Well, sounds like great growth. Congratulations on all your success in the company. Again, the wave is just beginning. Sounds like you guys built that big surfboard to ride the big wave, big wave riding going on here with AI. As we close out, take a minute to explain and the quick story, the bumper sticker for the company, the value proposition, and then any advice you'd like to share to aspiring AI folks out there, entrepreneurs, developers, companies who want to jump in, jump in the deep end of the pool, start to swim in fast. Quick summary of your pitch, bumper sticker. So I can talk about, so we've been focused a lot on talking to the executives and leaders. So I can talk about what leaders need to think about in terms of what's happening to take the advantage of this approach. The first and foremost thing is to accept and realize that there is a paradigm shift in AI with generative AI. We are not going to be anymore training one model for a task. There will be a foundational model that will help enable a lot of tasks, right? This will require different types of skills and engineering capabilities and customers will have to invest in upskilling their team. So that's the first part of it. The second part of it is basically understanding within their value chain. Where do they see the maximum value, maximum ROI or maximum impact of generative AI and start focusing on them? And the last one I would say is to find a champion within their organization because any big thing like generative AI applications will need a champion within the organization to move it forward and that's going to be a very, very important feature. It won't be hard to find a champion in the company that anyone who's got tech chops to go all over AI. I mean, I wish I was 25 again, Dustin. I mean, we love it. I mean, AI is the fountain of youth, you know? It's really exciting. Yes, definitely. For sure. We are very excited about what the world is going to look like in the next three years and five years. Our industry is going to adopt a technology and have a new way of working. So thanks for coming on theCUBE. You know, we're Chile's AI. We love AI. We're Maximals on this. We think it's going to be game change. We're seeing here already at Google Next and we're looking forward to checking in with you guys later. And here, what you're hearing from your customers, we'll come back on theCUBE another time and we'll check in with you. Thank you for coming on. So definitely. Thank you for inviting us and we look forward to conversation again. All right, good job. Well done. CUBE here, live. Get all the data, laying out the frameworks, his best practices, his new ways of doing things. This is a way that's going to change new values, going to create process change, create opportunities for entrepreneurs and innovators to navigate the complexity of the legal and compliance. That's going to be done very easy. It's going to be done in a way like this. AI is here to stay. We've got coverage here on theCUBE. Thanks for watching. We'll be right back.