Google Cloud Marketplace Marvels | Google Cloud & Elastic

Clips
News
More from Google Cloud Marketplace Marvels

Stephen Orban

VP of Migrations, ISVs & Marketplace

Google Cloud

Ken Exner

Chief Product Officer

Elastic

Partnership between Google Cloud and Elasticsearch helps customers solve big data problems with generative AI

When Google LLC announced new generative AI features for its cloud platform in April, the company provided users with an ability to use technology from Elasticsearch Inc. to leverage cross-network data for model training.The announcement highlighted how Google Cloud has been working with its partners to offer a one-stop shop for enterprise generative artificial intelligence development.“Every customer has a … data problem,” said Ken Exner (pictured, right), chief product officer at Elastic

play_circle_outline Google Cloud Marketplace growth and customer commitment

play_circle_outline Elastic's success with Marketplace sales

play_circle_outline Significant product integrations and technology innovations with GCP

play_circle_outline Unlocking Customer Value: Integrating Data-First Strategy, Generative AI, and New Integrations

play_circle_outline Customer benefits and success stories with Elastic, including Cisco

play_circle_outline The Evolution of Elastic on GCP: Maximizing Marketplace Growth for Seamless Customer Experience

Info
Transcript

Google Cloud & Elastic

Stephen Orban

VP of Migrations, ISVs & Marketplace Google Cloud

Ken Exner

Chief Product Officer Elastic

Stephen Orban, VP at Google Cloud, and Ken Exner, CPO at Elastic, discuss their successful partnership and integrations. Elastic's shift to the cloud and use of the Google Cloud Marketplace has led to increased business growth. Selling through the Marketplace reduced sales friction, resulting in 75% of their GCP cloud business coming from there. Elastic's integrations with Google Cloud services like Storage, BigQuery, and Pub/Sub have enabled customers to index data and gain insights in areas such as generative AI and security. They focus on real-time data, a... Read more

explore Keep Exploring

What are the two ways in which customers can satisfy their commitments to Google Cloud's Marketplace? add

What were some benefits of using a Marketplace sales motion for Elastic Cloud? add

What are some of the integrations Elastic has with Google Cloud Platform services? add

What are some common use cases where customers are leveraging the combination of Google's data processing and storage with Elastic's search and analytics and AI capabilities? add

What is an example of a company that uses GCP and Elastic for various search use cases, including text-based and semantic search, as well as empowering support engineers to answer cases more efficiently? add

What potential opportunities do you see for enhancing the presentation of Elastic and similar partners within the context of GCP in order to better serve the developer community and customers? add

bolt Powered by CUBE AI

Google Cloud & Elastic

search

>> Hello, welcome to this special CUBE presentation. I'm John Furrier, host of theCUBE here in our Palo Alto Studios for the Google Cloud Marketplace Marvel series. We've got two great guests here, CUBE alumni, Stephen Orban, vice President, Migrations ISVs, and Marketplace for Google Cloud, and Ken Exner, Chief Product Officer for Elastic. Ken, it's great to see you.

Ken Exner

>> Good to see you.

>> Stephen Google's, a lot of cloud momentum showing tons of momentum. We saw that at Google Next, how are you today thinking about the Marketplace? What's going on?

Stephen Orban

>> John, maybe before we go, I'll just give a couple of words on how Google thinks about the Marketplace and also thank Ken for being the first to go on this Marketplace Marvel series. Of course. John, thank you as always for hosting. It's a super exciting time for Google Cloud and an exciting time specifically for our Marketplace. As we continue to see significant growth for ISVs, the software companies we partner with, who really lean in and make Marketplace their default way to meet customers where they are on their journey with GCP. And as more and more large customers like Home Depot, Goldman Sachs and Mayo Clinic make significant moves to Google Cloud, our backlog of committed spend has grown to more than 70 billion. Now, customers have two ways to satisfy those commitments. One, by consuming Google Cloud products, or two, by consuming one of the thousands of ISVs on our Marketplace. And as we've added more and more capabilities for ISVs to find and reach customers, including selling alongside them like we do with Elastic, with our field teams, and we recently added support for customers to buy from the reseller of their choice, we're finding that customers actually migrate faster when they buy from the right ISVs at the right time. And ISVs can close larger deals faster and more often. But at the same time, growing with the Marketplace is just like many things in life and in business, you get what you put into it. So we started this series, and Ken thanks again for being the first to do this with us, really to shine a light on what some of the ISVs who do this really well, how they've done it both to help our customers understand why Elastic is such a great partner, and to help other ISVs achieve the same levels of success. So you'll hear from Ken on how Elastic has leaned in on all of the three necessary components of a successful partnership with us on Marketplace One, they build unique solutions that customers can only find on GCP. Two, they market it well to their customers and the right places of their product experience. And three, they engage in a deep co-sell relationship where our field teams work closely together to make sure we're solving our customer's toughest problems together.

>> Ken, how have you guys been able to grow your business with the Marketplace and increase your go-to-market speed? What's been the secret?

Ken Exner

>> Yeah, so Elastic began as self-managed software. And as we moved to the cloud to offer a managed version of our offering, of course Google was one of our CSPs that we wanted to grow with. And initially, we started a direct to customer sales motion where we would directly sell to customers. As we added Marketplace, we realized that there were a ton of benefits from Marketplace. So we started selling Marketplace as well as direct on GCP. But what we realized is that when customers were using a Marketplace sales motion, one, there was less friction in the process. They didn't have to go through procurement, they could immediately start using Elastic on GCP. So this meant less friction, this meant greater overall consumption of Elastic. So we started seeing that this was actually a better avenue for us. So we actually started promoting Marketplace-based sales of Elastic Cloud, because customers were going to spend more, they're going to have less friction. It was a better experience. So we did a few things. One is we changed our pricing, so it didn't advantage us. It would be the same whether you're coming through Marketplace or directly through Elastic. And the second is we changed our onboarding experience so that when we sign people up or when we log them in, we put GCP Marketplace right there next to our direct channel. So we're promoting that avenue just as much as our own. And today, I think 75% of our GCP cloud business comes from Marketplace. So it's been phenomenal for us. And I think that's why we've won Google Cloud Partner of the Year four times in a row. Four times in a row. Is that right, Stephen? Four times in a row?

Stephen Orban

>> Four times in a row. It's among the reasons, Ken, those are some definite best practices, but you've also got some awesome integrations with us as well.

>> It's quite a marvel, I must say. Glad we named it the Marvel series. No, in all seriousness, I think that's a great point. No conflict. Seamless buy-in. Great go-to-market benefits. That's awesome. Congratulations. Love the Marketplace formulas. It's like direct all about. So congratulations on that and congratulations on the award. Let's get into the product, the partnership between Google Cloud and Elastic, Stephen, you also mentioned up front, has yielded a lot of significant product integrations and technology innovations that's unique to GCP and also for Elastic Search, Observability Security. Can you guys share and discuss the integrations and what you guys have done together?

Ken Exner

>> So Elastic is used to search and index and find information and analytics and data. So that means that we need to support all the different data stores in GCP. So for a long time we've had integrations with Google Cloud Storage, with BigQuery, with Pub/Sub, with Dataflow, a bunch of integrations, connectors to GCP services so that customers can index that data and get value through Elastic. I think some of the more interesting stuff is what's happening lately though with generative AI. Where people who use Elastic for generative AI use it as a bridge to LLMs for their private data. They have a bunch of private data, they want to connect that to an LLM, and what they do is they use Elastic to do that. So if they're wanting to ground an LLM on private data, it's a search problem. They want to search for the most content and then pass it to an LLM. And what Elastic does is it integrates with Google in a couple of different places here. A typical workflow is you want to ingest data, you want to run inference on that data, store the embeddings in a Vector database, in this case, Elastic Search, and then run retrieval and then pass context to an LLM. Well, in the inference part of this, we not only support our own inference model, we support Vertex AI, embeddings creation services so that you can do that directly from the inference API in Elastic. And we support passing context to an LLM at the end of that workflow. In this case, Gemini Pro. So you can use Elastic together with the best generative AI at Google to create a complete workflow for grounding and building an LLM-based application on top of Google and Elastic.

>> What about security? That's a big point. You've just brought that up. People have their own data. That's been the number one focus for enterprise AI is going beyond the chatbot, but really making the private data valuable. Does Vertex help there?

Ken Exner

>> Well, for a number of reasons. One is a typical SOC, a typical security operations center is going to have a number of different tools for threat hunting. And a very popular combination is using Elastic together with Chronicle, together with Mandiant. Chronicle out-of-the-box integrates with Elastic. Elastic out-of-the-box integrates with Mandiant Threat Intelligence. So modern-day SOC is going to have the best of those combinations for doing threat hunting. But what we've done in Elastic is we've also built various AI capabilities on top of Gemini and Vertex AI that help to automate some of the capabilities of the typical modern-day SOCs. And not just AI assistance, but entire capabilities of automated threat detection. So we have a new feature called Attack Discovery, for example, that is built on top of the power of Gemini, that allows a typical security analysts to look at all the different alerts that they're getting and map that to an attack chain. So the modern-day SOC is improved. The information that they have is automated by the combination of Elastic and Google.

>> Talk about the observability aspect. People want real-time data in real-time. It's been a big part of making generative AI work is if you don't have the data that AI doesn't really work. Well talk about the real-time aspect for observability.

Ken Exner

>> Well, of course. So if you want a real-time, when you're thinking about observability, you need to know about things the second they happen. So if you have some kind of operational issue, you want to get alerted very quickly and you want to have anomaly detection that runs to find things that you can't necessarily predefine as an alert. So the combination of predefined alerts together with anomaly detection based on generative AI and ML models will allow the operations analyst to get the best of both worlds, things that they define as well as things that are driven by AI to help them be alerted of things as they happen. And then because of the integrations that we have with Gemini and the LLMs, we can start automating the remediation flow. So you can automate the runbooks that are used by operators to remediate issues. And this is because we have the context of their operating environment, we have information about what went wrong, and we have the power of LLMs, in this case, Gemini, to automate the production of a runbook for how to remediate an issue.

>> Enterprise is looking to unlock value with data. That's the key part of the piece here. You mentioned security, that's obvious, that's mission-critical. But data first strategy is becoming something important and developers are a big part of it. Can you guys talk about the developer impact of the integrations? Because at the end of the day, developers want to go faster, they want faster applications, the search is critical for them. Having scalability is all going to be key. How do you guys talk to the developers?

Stephen Orban

>> Ken, you want to start?

Ken Exner

>> Sure. So yeah, Elastic has always been developer first. We love developers. We were built as a platform for developers. Our APIs for search and for analytics have been loved by developers for years. We have conferences where we bring developers together to talk about the things that they build on top of Elastic. The thing about Elastic is it's not just a search engine, it's a search AI platform. And what I mean by that is it's a platform for building applications. If you think about, we started off as a way for you to integrate search into a website or into a product, but as people started realizing the power of search and search AI, they realized that you can build financial analytics applications on top of it. They realize that you could build ride-sharing applications on top of it. So when you search for a car, you're using search. When you do matchmaking searching sites, any of these things are actually search applications. So people are able to build a completely new novel applications based on the power of search and search AI.

Stephen Orban

>> The thing I would add to that, Ken is we share at Google Cloud a love for developers and specifically trying to meet them where they are. And I think one of the reasons that we find developers like Elastic on Google Cloud so much is because of how many unique and expansive integrations you've done. I mean, you just talked about a couple where you're integrating with Vertex, where Elastic can be a data source where the search and observability helps across our security command center and our Mandiant products. You've got a number of integrations into BigQuery. These are all reasons why developers choose Google Cloud. And the fact that Elastic, despite also being a multi-cloud product, you offer your services anywhere the customer wants, just like we want to have any ISV that the customer wants on our service, you're still making a bunch of those unique integrations that provides the developer faster time to product and time to value in a way that really makes both Elastic and Google Cloud shine together in unique ways.

>> That's awesome, guys. So congratulations on the partner show. I got to ask the question that everyone wants to know is what's the unique value from this combined relationship for your customers?

Ken Exner

>> I'll start. So every customer has a big data problem. Every customer that we have in common has tons of data that's being emitted by their operating environment, that's being emitted by their applications. And the combination of Google's data processing and storage together with Elastic's search and analytics and AI capabilities create a powerful way to sift through all that data to get to insights and to get to actionable, to get to things that they can action on. So I think every single one of these use cases that we have in common is a big data problem where a customer wants to get through data based on stuff that's being hosted and processed by GCP and analyzed by Elastic. But there are some very common use cases that we see across our customers who are building search applications on top of GCP or building AI applications. I think these are the ones I'm really excited about where customers are starting to leverage Elastic to do semantic search on top of GCP, or they're using Elastic to do vector search and start grounding LLMs on their private data. It's really exciting to see how many customers are starting to take advantage of these capabilities that the two companies provide in common in these joint solutions.

>> Stephen, Elastic's been a great product. It's been from the beginning, go back a decade plus, I mean, it was really the product, if you were doing anything in hyperscale, developing Elastic was the key product. Data's now more important than ever, and integrations we were just talking about earlier has become a big part of the generative AI wave. People need to integrate models, foundation models, also data sets, unstructured and hybrid. Since a big part of the new equation continues to be important, but it's changing. You've got neural networks, as you mentioned, vector databases and tokens. RAG is a popular application. You're starting to see a lot more computer vision. A lot more new things. It's a net new category emerging. So how does that change the integrations, and has Marketplace evolve and Elastic two is in the middle of it?

Stephen Orban

>> Yeah, so it's a great question, John. And I think just to double down on the uniqueness that Elastic provides, there's no other ISV that I can think of, no other software company who handles all of search observability and security all in one wrapped up offering that is also uniquely integrated into so many of our different service lines. And when you talk about data and what the impact of that is into customer's gen AI pursuits, look, the models that we make available are great, but they're currently trained on publicly available data. So the customers who are doing unique things with their gen AI capabilities, whether they're putting chatbots that are trained on their own, customer service data or whatever the case may be, they need to be either tuned or augmented using let's say retrieval augmented generation with data that they own and their own private corpuses of data. And oftentimes, customers are already using Elastic as the storage layer for that corpus of data. So some of the things that we're doing around Vertex are really helping accelerate some of our customers, gen AI, both proof of concepts and then ultimately getting them to production faster because we're able to collectively meet the customers where they are. Ken and the Elastic team with the search and the data storage, and then Google Cloud with the models, whether it be Gemini and our platform Vertex together is like peanut butter and jelly.

>> Ken, you got to love the market. I mean, Elastic's been in the search because you pointed out clearly, but now the whole RAG or retrieval augmentation general, that's essentially search that's retrieving. So now you have generative AI, which needs a scalable search function basically. So I'm oversimplifying it, but the whole world has gone a search as a layer. This is now standard for all gen AI applications. So you guys got to be pretty excited over there and thinking, well, the data's also there too. You had the data in your platform. Well, you guys are well positioned for this big wave.

Ken Exner

>> Yeah, we are totally excited about the opportunity ahead of us and what customers can do with Elastic as a bridge to generative AI, we actually found ourselves in the position to be a part of the modern generative AI tech stack because of investments that we made a long time ago. So we of course started as lexical search or text-based search, but as customers wanted to start doing semantic search, natural language, question answering or image search, we needed to start supporting vector embedding. So being able to generate and query vectors, and that sort of positioned us to be a solution for RAG these days. We actually supported vector embeddings and vector search in 2019. So more than five years ago, we were starting to do this already. Today customers have been indexing their data with Elastic, they already work with us, and they can start passing context to an LLM with the solution they already have.

>> You guys nailed the developer experience, not to pat ourselves on the back of theCUBE here, but we've been a customer since 2012 when we started indexing all of our content. Elastic was our first solution. We never left. So you guys got a great product. Anyone doing horizontally, scalable cloud data pretty much has used Elastic or uses it. So really well positioned. Let's get into the customer benefits now because now the proof is who's using the product? What's the new use cases? Can you share the customer success story? You guys have Cisco as a customer success story. Take us through that.

Ken Exner

>> Yeah, Steve and I were talking about Cisco earlier. Great example. It's a customer of GCP and Elastic. They run all their infrastructure on GCP. They use us actually for a number of different search use cases. Cisco.com uses us for their search on their website. They also use us for their internal search on their internet, sort of a workplace search use case. And they also use us as part of their support center in empowering search for support engineers who are trying to answer cases. I think the interesting thing about what they've done is that they've grown with us. So they started off doing text-based search, and then as we introduced semantic search, they added semantic search so they could do natural language processing. Then as models came about or they could start automating the creation of answers, they started adopting that from Elastic. So they do natural language question answering. And they actually today implement something, a form of hybrid search where they do a combination of different techniques where they can suggest answers or point people to specific pages where they can get more information. They've rolled this out across all the different search use cases and they've had phenomenal results. I hear that they believe that they're saving about 5,000 support engineer hours every month because customers are able to find information themselves or the support engineers are able to find information faster. So they've had phenomenal results using Elastic to power search, but using us together with Google to provide generative AI-based answers.

>> So they bring the data sets together? Were they siloed? Did they unify them? What was the key problem and then did it match up on the performance side? Can you guys talk about that specific aspect of it?

Ken Exner

>> Yeah, they have a number of different silos. If you think about an internal knowledge base, they have various wikis, they have various different tools that they use for ticketing and for all kinds of different silos of information internally. But also in terms of support, the support workers need to reference data that lives in dozens if not hundreds of different places. So what we do together with Google is index all that information, combine it with the power of LLMs to help provide summarized answers and then point them to the reference documents. So we ground the answers based on the documents they have that are distributed across the entire Cisco network, and then provide them a simple answer that can then point to the underlying reference documents. So it's a powerful use case that has saved them a ton of time and resources.

>> How's the performance on the queries latency? Can you share some stats on performance?

Ken Exner

>> Well you know Elastic, we are known for our speed and performance.

>> Yes, that's for sure.

Ken Exner

>> If you're talking about search these days, it has to be fast enough to do auto-completion, right? So what they've done, they've actually implemented. So as you're typing, it starts auto-complete answers in this hybrid search fashion. So it's fast enough where it is doing it as auto-complete.

>> And that is, by the way, generative AI is all about completing the next sentence, right? That's runtime. That's the kind of speed you need. Awesome. Stephen, you want to lay way in?

Stephen Orban

>> Yeah, the only thing I would add to what Ken said is just he kind of walked through it, it was like a growth motion where Cisco started with one or two use cases and then expanded into adjacent areas, because they continued to see increasing value, performance, speed, et cetera. It also lends itself really well, that type of motion, to what we try to do with our marketplace where there's a very easy sort of low-cost way to get started. That's frictionless. That just shows up on Cisco's bill. It automatically complements whatever they're doing, what they were already doing with Google Cloud, where they host the rest of the adjacent workloads. And then over time, it's a great forcing function for our account teams to really get to know one another better, which they did in this case and grew with Cisco over a longer period of time. So I think in addition to it being a great technical win and driving a bunch of product value for Cisco is just a great way for our teams to work together and really understand what Cisco's pain points were so that we could both bring our respective best of breed technology and capabilities to the table.

>> What I love about the cloud next gen, and certainly with generative AI on the corner, data is the center of the value proposition. It's got to be large scale, horizontally scalable, vertically specialized. The application is going to get the developers in line with that. Data's now moving up the stack big time. So I want to ask you guys both to wrap up here, what's special about integrations with Cloud and ISVs? Ken, you're the customer in the Marketplace. You're moving a lot of product, sales are up. Stephen, you're enabling a marketplace. What makes this time so special for the folks watching out there saying, "Hey, you know what can I do different? I just don't want to lift and shift and be in a catalog and move product. What can I do to make my business better?" What do you guys offer Stephen? And then Ken, how would you talk to a colleague and saying, Hey, this is a good deal?

Stephen Orban

>> Yeah, John, thanks for the question. I think it's a super exciting time to be in tech and working with Elastic in particular because generative AI is going to make way to so many use cases we haven't even thought of yet, but also automate a lot of existing things today, whether it be customer service or summarizing documents or generating code or images or marketing campaigns. And I think the unique set of capabilities that Elastic has built with us over the years, not just in our generative AI platform with Vertex and Gemini like Ken was talking about, but also access to BigQuery as a data store and then securing things through Mandiant and the like, really positioned as well, to help customers not just identify those use cases, but implement them very quickly.

>> Ken, what's your take as you look at the product opportunities to innovate and other peers out there thinking and scratching their head trying to put this puzzle together as well, what would make them a marvel?

Ken Exner

>> Well, I think one of the things that we've been seeing in the industry for a while is how do you deal with all this data? And typically, people have been trying to reduce the amount of data that they have. And I think what's different these days is that it's okay to have all that data, if you can search through it and parse it and then pass it to an LLM and have an LLM help you get answers out of it and process it. Actually the more data the better. And that's a new thing that we are realizing. When you're thinking about if you're securing your enterprise, for example, you don't want to reduce the amount of data and then try to process that. Just take all the data and work with Elastic to find the right context out of that data and then pass it to an LLM. So I think the situation has changed where the more data, the better, not the worse. It's okay to have a lot of data and just use the combination of GCP and Elastic to help you get answers and actionable inside that of that.

>> I think that's a great point, and I think the neural network infrastructure, you mentioned vectors before, will allow you to get billions and billions of more data points and retrieve that in real time. You got to have the cloud capabilities, you got to have those agents, you got to have the APIs, and really it's a whole unique time, Stephen, this is a unique. This isn't just host on the cloud. This is what Ken was getting at, is that more data is the better because you feed the AI, it gets more intelligent. That next level is all about a new kind of infrastructure.

Stephen Orban

>> That's right. And then you need the cloud to be able to scale all of that, both that data and the computing resources around it. So it's a super exciting time.

>> Stephen and Ken, thanks for coming on this series. What's next? Ken, what's next for you? Steve, what's next for you? Ken, we'll start with you. What's next?

Ken Exner

>> Well, we're continuing to work on relevance at Elastic. I think so much of the RAG discussion has been around performance and capabilities around RAG. I think the secret is actually relevance, because if you want to pass context to an LLM, you want the most relevant information because that generative AI application is going to answer something. It's going to give you one answer. So you want that one answer to be right, which means that you have to have the most relevant grounding. So we're working on a number of techniques, whether it's re-ranking, whether it's query rewriting to provide additional capabilities for developers to tune relevance, because relevance is what matters when you're trying to find the best information. So I look forward to coming back and talk about what we're doing there.

>> I can't wait. Reasoning, relevance, value. Stephen, what's next for you in the Marketplace? What's going on?

Stephen Orban

>> So as fast as our Marketplace has grown over the last couple of years and as fast as it's grown with Elastic specifically, I still think we have an opportunity to make sure that we're presenting Elastic and partners like them in the contextually relevant spaces across GCP so that we can make sure we're presenting them to the developer community and to the customer when it's contextually relevant for them. So Ken talked about some of the integrations that we've done already, but I think there's more that we can do. Specifically to make sure when a customer is starting to build or train a model, they could discover an Elastic data source that they're already running in GCP without necessarily having to leave the experience that they're in. We've started that integration already by Elastic working with us and training our Gemini Code Assist with Elastic specific data, so our models are smart on how developers use Elastic, but there's a lot more that we can and will be doing as the future unfolds.

>> Stephen, Ken, thanks for coming on this Marvel series. Cloud Marketplace Marvels. Looking forward to it. Is good a great time right now to innovate with integrations and data. Again, next level, just scratching the surface. Thanks for coming on.

Ken Exner

>> Thanks, John.

Stephen Orban

>> Thanks.

>> Okay, Google Cloud Marketplace Marvels series. I'm John Furrier, host of theCUBE. Thanks for watching.