Google Cloud Marketplace Marvels | Google Cloud & Databricks

Clips
More from Google Cloud Marketplace Marvels

Stephen Orban

VP of Migrations, ISVs & Marketplace

Google Cloud

Shanku Niyogi

VP, Product Management

Databricks

play_circle_outline Databricks' mission to democratize data and AI

play_circle_outline Commercial model aligned with Google Cloud's consumption-based model

play_circle_outline Deep integration with Google Cloud for optimal performance

play_circle_outline Real-world example: Uplight utilizing Databricks on Google Cloud for energy consumption insights

play_circle_outline Achieving Global Success: Databricks Integration, Strategy, and Market Fit Across Regions

Info
Transcript

Google Cloud & Databricks

Stephen Orban

VP of Migrations, ISVs & Marketplace Google Cloud

Shanku Niyogi

VP, Product Management Databricks

Stephen Orban, VP of migrations, ISVs, and marketplace at Google Cloud, and Shanku Niyogi, VP of product management at Databricks, join theCUBE to discuss the evolving landscape of cloud computing and AI. Hosted by John Furrier in Palo Alto, the conversation delves into the integration of Databricks with Google Cloud Marketplace, aiming to democratize data and artificial intelligence. Understand the significance of the unified data intelligence platform and how it empowers various stakeholders, from developers to business users, to extract value from data eff... Read more

explore Keep Exploring

What is Databricks' mission and how does their data intelligence platform integrate with the Google Cloud Platform? add

What are some advantages of using Google Cloud Marketplace to access Databricks for customers? add

What are the unique aspects of Databricks and how does Google tie into that in terms of integrations? add

What company has been able to achieve incredible energy consumption savings by using Databricks on Google Cloud? add

What are some key factors contributing to the success of Databricks in various international markets in the past year? add

bolt Powered by CUBE AI

Google Cloud & Databricks

search

>> Hello everyone. Welcome to theCUBE. I'm John Furrier, your host. We are here in our Palo Alto studio. And welcome to the Google Cloud Marvel series where we talk to the leaders of the industry leveraging Google Cloud and the Google Cloud Marketplace. This conversation's on the evolving cloud and AI landscape. We'll dive into the partnership and how leading technology companies like Databricks are driving innovation together. I'm joined by Stephen Orban, VP of migrations, ISVs, and marketplace at Google Cloud, and Shanku Niyogi, VP of product management, Databricks. Guys, welcome to theCUBE, and looking forward to a great conversation.

>> Thank you, John.

Stephen Orban

>> John, thanks for having us. Shanku, thanks for doing this.

>> Shanku, Databricks obviously continues to be on a tear. Ali continues to announce more and more success. And open source and what you guys have built is continuing to thrive and enabling this new generation as well as the existing transformation in cloud and now the AI market. Google Cloud has been a key part of that. Want to get into it. Talk about the relationship with Google Cloud Marketplace because I think that has really been a key driver for some of the successes on the customer side. Take a minute to explain the Databricks approach and your relationship to Google Cloud Marketplace.

>> At Databricks, our mission is to democratize data and AI. And our data intelligence platform is a unified platform that brings together data analytics and AI with a unified way to govern, and this concept of data intelligence that really builds AI integrated with data so that really anyone can get value out of data. Whether you're a developer building an AI application or an analyst doing insights from data or just a business user who's using natural language to get knowledge from data. I think the key value of integrating that with Google is to really use the power of the Google Cloud Platform to be able to do that at scale in a way that's really industry-leading. And a key part of that for us has been this integration with the Marketplace. I think this month about four years ago is when we launched our integration, and since then it's been a tremendous partnership. We've got now over a thousand customers. And pretty much every one of those, I think 98% or something of those customer journey start in the Marketplace. And we have a very common business model of being able to be powered by consumption, having convenient pay-as-you-go options, and that allows anyone on Google Cloud to just get started and grow really limitlessly on Databricks.

>> I'd love to have you explain, and Stephen if you can chime in too. Databricks really hit really this tailwind of what you guys were doing with data. And we've been, again, following the evolution. But we are in this modern data and AI ecosystem, how has Databricks really taken advantage of that? And Stephen, how does that tie in to what Google's offering? Because we are seeing a new modern approach from how open source has changed radically on the software development side, but how data's being used to power AI. We have a modern data and AI ecosystem developing. Shanku, talk about the Databricks approach in this modern era. And Stephen, if you don't mind, chime in over the top.

>> Yeah. It's amazing. I think Databricks and Google, I've been at both places, we share a lot of common approaches to this. We believe in the value that data brings. We believe in the power of AI. We're both deep believers in the power of cloud-native technologies to be able to scale these kinds of platforms to get business value. And we both believe in open architectures and open source between things like Kubernetes at Google and Apache Spark, which really started Databricks' journey. So this has been a really great partnership. I think part of this is not just being able to leverage Google Cloud but deeply integrate our ecosystem so that you can use our products together as a customer.

>> Stephen, modern data and AI ecosystems, big part of your strategy at the Google Cloud Marketplace. Talk about your perspective.

Stephen Orban

>> Yeah. I mean Shanku really nailed it. There's so much we have in common with our approach. In fact, I'd call Databricks and Google Cloud are like peanut butter and jelly. Or if you have more of a sweet tooth, maybe peanut butter and chocolate that go together so well. But the reason we have a marketplace, the reason we have an ISV ecosystem is we know that we need to meet our customers where they are. And no matter how great our technology is, customers are always going to have an insatiable appetite for more and different tools at the right price-performance point from their perspective. And you look at Databricks and they've been just on a tear or on a roll for the last bunch of years. They've had great product market fit for a long time and their product continues to get better at a rapid clip. And they've continued to acquire great companies that are making their product better and more strong. And customers really appreciate the way they're able to unshackle the silos of their data across their organization and now start to apply artificial intelligence to it. And I think, John, you nailed it. There's a lot of tailwind now in the market with customers really trying to figure out not just what their AI strategy is but to do that they have to have good, clean, available, accessible, well-governed data to power all that, and Databricks really checks all those boxes.

>> Talk about the Databricks growth. Obviously, we've been seeing... You're in the VP and the product side so you get... There's always new products coming out. I've never seen more product launches coming faster. Accelerated market we're in right now. How has Databricks evolved this go-to-market strategy to align with the rapid growth of these AI-driven workloads? Not just the workloads that are going into production, but a lot of people working on setting up that tsunami of more AI-driven workloads. As more products come out, how are you aligning that go-to-market strategy with the growth?

>> Yeah, I think our growth is driven by this singular mission around democratizing data and AI and making it accessible to all. If you start from our inception, building a business on Apache Spark that made working with large amounts of data incredibly easy to do. Growing from there to where we brought SQL warehousing and this concept of our open lake house, that really changed the way to think about data warehousing and big data together. And then we brought in Unity Catalog, which was really to help unlock data at scale across the company in a way that was secure and governed in a single way and bringing in with Google's data state together in a way that's deeply integrated and governed. And then adding this concept of data intelligence where we can put AI models to work on top of that. Ultimately each of those things is expanding our footprint to new classes of customers, new segments of users, with the goal of really helping every organization unlock their data at every level. So we've gone from those core central data teams to where Databricks is now a business value proposition for anyone in the company and really going up to that CIO, CMO, CXO level. And that's great because now we can look at a platform level how Google and Databricks are really integrating to deliver value across the company.

>> So my follow-up on that is: How are you guys taking advantage of the Google Cloud Marketplace, and how are customers leveraging it? Obviously, consumption is great on the customer side. How are you guys using it and what's the customer's consumption look like?

>> I think the nice thing about Marketplace is it lets us really align our go-to-market motions in a global way. I think most of our customers are getting Google Cloud through the Marketplace. The fact that the consumptive business model is aligned and customers can use their EA commit to be able to grow, and Databricks gives them that limitless value to scale. And that goes really nicely with the platform value proposition. The other piece of that is I think Databricks... Getting to scale at Databricks starts with getting started. And so the fact that we built this super simple experience where you can just go to the Google marketplace. If you're a data analyst or a business user, you no longer have to be a cloud administrator, you no longer have to have a central cloud account. You can just get started with Databricks in just a few minutes and start to get that initial value. That is huge in terms of our ability to then grow from there.

>> Of course, AI is going to make it even better. Stephen, Google Cloud's perspective, how is the Databricks integration unique and what are you guys doing on your side to enhance their value proposition?

Stephen Orban

>> If I could just double down on some of the things that Shanku just said there. In addition to having a great product, they have a very compelling commercial model when it comes to customers and particularly developers who just want to get started as he pointed out. And they have this concept of a DBU that they sell where you can get started with just one. It doesn't have to be a huge commitment to hundreds of thousands of these things. You can just get going really easily. So that has just a great parallel effect with how cloud providers go to market also. We sell machines and storage by the hour, in some cases, even by the minute. And then customers as they graduate and use more can of course make commitments for larger discounts. But most of the promise is that you can get started quite easily and friction-free. And I think Databricks has just done a world-class job of taking that into account, working with us to design their marketplace listing in such a way that it's very accessible and easy for customers to get started. And then the proof is in the pudding. The customers start to use the product at a very low fee for a small bit of consumption. And then as they start to spread that out, or let's say one business unit uses it and starts to share data with another business unit, and then that business unit starts to get some value out of it, they can graduate into higher tiers of consumption. And then, of course, because they're doing it on the Marketplace, not only is the procurement streamlined and easy and all of that shows up on their Google Cloud invoice. But for customers who have large commitments with us, they're able to draw down on those commitments for their Databricks spend as well. So again, it just comes back to this peanut butter jelly and a high level of alignment between not just our product but our go-to-market motion.

>> That's a great point.

>> I'll add one other thing to that. I think in alignment with that, the fact that the product is so cloud native. We use the same security model, the same compute model, and there's these integrations. Using open technologies between Google products and Databricks means that from a customer perspective, they can see that their cloud estate is one. And they can do things like combine our AI model together or Vertex AI together with MLflow to really operationalize large AI or be able to combine our analytics tools together with BigQuery to get fast insights. And every one of those impediments in the business model and in the product we've tried to address to really optimize for customers.

>> Shanku, that's a great point. And Stephen, great point on the commercial license. Every new wave has a pricing change. Perpetual license in the old software, cloud, software as a service, now you've got usage and all kinds of new ways to do business. So great alignment there. The integrations become a big thing too. So now you've got business model innovation on the commercial side. Shanku, you brought up those integration points. This becomes a key part of how we see people succeeding in this new AI ecosystem. Connections aren't just an API call, there's more integration seems to be on the successful players. Talk about the integration specifically. Okay, marketplace, great commercial value. Check. Now integrations become where the scalability comes in because once you start, you can start any entry point. Talk about the integrations. What makes Databricks unique and how does Google tie into that? If you both can talk about that, I'd like to get into the integrations more.

>> I mean the thing to start with here is, Google Cloud is a great place to run Databricks. When we went and launched that integration four years ago, we deeply integrated with Google's containerized platform. This was actually the first fully containerized way to run Databricks. And that meant that you could take open-source Spark workloads and run them 50X faster on Google Cloud, on Databricks, which is immediately unlocking an immense amount of value and scalability for customers. Even just this month, we launched our new serverless general compute platform, which means that whether you're doing interactive data science or data engineering, you no longer have to have all the knobs to have to manage in order to grow your workloads. And we've done that by really integrating our serverless technology in a deep way with Google's platform. At the data layer, what's nice is open source technologies like Delta Lake are enabling us to integrate BigQuery together with Databricks. So by having Delta support in BigQuery, it really enables Databricks' customers on Google Cloud to be able to seamlessly go between BigQuery and Databricks to do their analytics. And then finally, I think AI is... Everyone's trying to get to production on AI, and I think a huge part of that is being able to... It's a very fluid space and you want to be able to mix and match technologies. And what's nice is you can use the best of what's on Google together with operationalization tools in Mosaic AI, things like MLflow, things like our assistants to be able to really go from idea to a viable production very easily.

>> Awesome. Stephen, you want to weigh in? This has been a key part of your strategy and value.

Stephen Orban

>> Yeah. So for those who are paying close attention, Shanku just mentioned a whole bunch of unique to Google services that Databricks actually integrates with. And I think it's a huge testament to the technical strength of this partnership where Databricks took the approach of actually taking advantage of all of the unique capabilities that Google Cloud has so that Databricks is actually highly optimized to run really, really well on Google Cloud. Look, I manage the relationships and the partnerships with all of the ISV partners that we have across our ecosystem. And one approach that some ISVs take is to just use the lowest common denominator tech across the cloud providers because it's just easier for them to manage or so they assume. But the problem is that that's not unlocking unique value for customers. So our customers now know that when they pick to run Databricks on Google Cloud, they're going to get the best possible experience on Google Cloud because of the way Shanku and his team are taking advantage of GKE and GCS and BigQuery and some of our unique services. And likewise, our field teams, our sales teams that are facing off with customers every day know that that's something that they can be confident about with our customers because of all those unique integrations. Again, it's a world-class approach that they've taken to a world-class product and partnership.

>> That's great stuff. And that's consistent with the industry. I'm seeing some of the big advances where people actually going lower-level service integration as well as higher-level services in the cloud. So a combination of both is great approach to unlock that value. Guys, let's talk about the customer real-world use cases. Love the story. Obviously, commercial, integrations, technical, better together. Peanut butter and jelly as Stephen you said. But let's get into some real-world examples of customers that have benefited from the Databricks-Google Cloud solution.

>> In just four years, we've added over a thousand customers now. So everyone from Fortune 500 Global Enterprises to innovative AI startups are running on Databricks on Google Cloud. As a Californian, I want to pick one that's near and dear to me, and that is this company called Uplight, which is doing incredible things with taking insights from data to really help energy consumption thinking about the climate footprint of that. Uplight really believes in this idea of taking data, gathering insights from that, and sharing that with consumers so that consumers can make better decisions to use energy more responsibly. And this is an immensely difficult problem. They've got something like 60 million endpoints, everything from thermostats to EV chargers that are generating data for them to be able to process and get insights from. And previously they were doing that with open-source Spark workloads. And being able to scale on that just became an immense problem. So they were able to use Google Cloud Marketplace to get started with Databricks on Google Cloud. They now have 30,000 plus I think Spark jobs. They're adding new applications and new insights every day. And it's already shown immense results. In 2022, those of us who were in California, there was massive heat wave. There were threats of power being shut down every day. They were able to take those insights and have that result in something like 1200 megawatts of energy savings that really helped companies to be able to avoid those rolling blackouts. So that kind of immense impact from being able to use AI together with data is amazing, and that's been powered by Databricks and GCP. And they've done some really interesting things there. Some of those integrations that I mentioned, being able to combine BigQuery together with the rest of their data stack, being able to use a whole bunch of GCP services together with Databricks, using MLflow to really manage their Vertex AI workflows so that they can optimize those AI models. That's been a great success. The other nice thing is using GCP and Databricks together means they can get to value quicker. As a lifetime developer, one of my favorite stories out of that is using serverless. Just because they were able to iterate on their value quickly because they had things like quicker startup times, they were able to get new applications to production 50% faster. That's amazing because in a fast-moving landscape, that ability to turn something from idea to something that's getting results in production is so valuable.

>> That's a great example. Stephen, summarize what were the key factors in their success. And how can other organizations learn from that example because that ties together the better together story? What was the keys to success and what can other companies learn from that?

Stephen Orban

>> Yeah, I really love that use case for a whole bunch of reasons. One, because it was for a great cause. And I think helping people understand climate and its impact better will continue to be a hot topic for some time to come, and it was great for the state of California. Just to maybe mirror that a little bit. Shanku is a Californian, I'm a New Yorker and some of the largest banks in the world operate in New York, and I spend a lot of time with them and have over the last couple of decades. And I know many of them are looking at... Not just looking at but using Databricks in a material ways to optimize how they do trade settlement or manage capital markets or do fraud detection and insurance, et cetera. Just in the last year alone, I visited Latin America, the UK, France, Germany, Singapore, Japan. And in every single one of those theaters, I've talked to customers locally in the region who is either evaluating or using Databricks in a material way to power some of those capabilities that Shanku did such a great job of articulating over the course of the last 20 minutes. It's the technical integrations, it's the go-to-market motion, being very consistent with how we go to market with cloud. It's obviously the great product market fit and just being super accessible to customers across the globe. There's a lot of great wins behind us, but really the best is yet to come.

>> Yeah, that's great.

>> I love that delivering peanut butter and jelly to both coasts in World War III.

>> It's global. East Coast, West Coast, it's a global opportunity. Guys, thanks so much for the Marvel series. Shanku, thanks for coming on. Looking ahead, guys, what's next? How do you see the customer expanding their use of Databricks and Google Cloud with their AI and data needs as they continue to evolve? Like you said, it's just the beginning. What's next? More integration? More fun? Take us through.

>> We're just getting started. The Gen.AI cyclone, if you will, hit only a couple of years ago, and I think we're all still learning as an industry how to get value from our data. And I think having a platform that is highly capable all the way from the infrastructure up and has these really tight integrations and enables these customers, I think we're excited about continuing this partnership.

>> Awesome.

Stephen Orban

>> Just getting started, that's right. Best is yet to come. Very excited for the serverless offering also that Databricks recently brought to Google Cloud. And really looking forward to see what customers build with that.

>> Cyclone of AI infrastructure advancements, AI development and software advances, and data in the middle. It's really good. Great sandwich right there. So guys, thanks so much for coming on theCUBE for the Marvel series. Really appreciate it. Thank you Shanku. And thank you, Stephen.

>> Thank you, John. Thank you, Stephen.

Stephen Orban

>> Thank you, John.

>> Okay. I'm John Furrier here in the Palo Alto studios here in California, bringing all the action. The Google Cloud Marvel series. Thanks for watching.