Google Cloud Marketplace Marvels | Google Cloud & Glean

Clips
News
More from Google Cloud Marketplace Marvels

Stephen Orban

VP of Migrations, ISVs & Marketplace

Google Cloud

Tamar Yehoshua

President, Product and Technology

Glean

Glean leverages Google Cloud for data silo solutions and connecting with key enterprise customers

Glean Technologies Inc. is a rising star in conversational search, harnessing the power of Google Cloud’s infrastructure to break down data silos and amplify its marketing support.Glean is an artificial intelligence platform that connects and understands a customer’s data across sources such as email, Google Drive, Slack, Teams, Jira, Confluence and ServiceNow so that practitioners across an organization can find the answers they need to do their work

play_circle_outline Growth and success of ISVs on Google Cloud Marketplace

play_circle_outline Glean's unique AI platform for enterprise search and knowledge retrieval

play_circle_outline Maximizing Market Reach: Glean's Collaboration with Google Cloud for Enhanced Integration and Sales Initiatives

play_circle_outline Partnership between Glean and Google in providing secure and personalized enterprise solutions.

play_circle_outline Importance of breaking down data silos and leveraging AI for digital transformation

play_circle_outline Excitement for future growth and advancements with Glean in the marketplace

Info
Transcript

Google Cloud & Glean

Stephen Orban

VP of Migrations, ISVs & Marketplace Google Cloud

Tamar Yehoshua

President, Product and Technology Glean

Stephen Orban, vice president of Migrations, ISVs, and Marketplace at Google Cloud, and Tamar Yehoshua, president of products and technologies at Glean, discuss Glean's success in the Google Cloud Marketplace. Glean is a work AI platform that breaks down data silos, offering standard enterprise search and a ChatGPT-like interface for natural language queries. Users can build knowledge graphs and productivity apps using APIs. Glean's success is due to unique solutions, effective marketing, and deep co-sell relationships with Google Cloud, resulting in increase... Read more

explore Keep Exploring

What are the key factors that have contributed to the successful partnership between Google Cloud and ISVs like Glean on Marketplace? add

What does Glean do first? add

What are the advantages of hosting Glean inside a customer's GCP project? add

What unique value does the combined Google Glean solution provide to customers? add

What technology is being used to help solve customers' hardest technical problems and how is it being implemented? add

What are the speaker's thoughts on their partnership with Glean and the potential for future growth and success? add

bolt Powered by CUBE AI

Google Cloud & Glean

search

>> Hello everyone. Welcome to theCUBE series here in the studios of theCUBE for the Google Cloud Marketplace Marvels Series. I'm joined with Stephen Orban, vice president of Migrations, ISVs and Marketplace with Google Cloud. And Tamar Yehoshua, who's the president of products and technologies at Glean. Great to see you.

Tamar Yehoshua

>> Thanks so much for having-

Stephen Orban

>> John, thanks for having us.

>> I'm looking forward to this segment. Glean's been really doing great, so congratulations to the team there. Stephen, this series has really been about your customers, partners and the services you guys are doing, just the magic and the marvels that have been happening with Google. And again, your momentum continues to thunder away, congratulations. This series is super important for people to watch, to get to learn and get informed.

Stephen Orban

>> Yeah. So first of all, John, thanks so much for having us and big thanks to Tamar and the Glean team for really being on this journey with us. And it's been a super exciting journey and a really exciting time for Google Cloud, an exciting time for our marketplace really as we continue to see significant growth for our ISVs, our software partners who really lean in and make Marketplace their default way to meet customers where they are, on Google Cloud. And as we see more and more large customers make significant moves to Google Cloud for their digital and AI transformations, our backlog of committed spend, the commitments that those customers have made to move to Google has continued to grow with it. And customers really have two ways to satisfy those commitments. First by consuming Google Cloud products, or by consuming one of the thousands of software companies and the ISVs on our marketplace. And as we've added more and more capabilities for ISVs to find and reach customers, including the fact that we basically sell alongside them with our field teams. So our field teams work hand in hand together to really help customers understand the ISV solution and why it works well on Google Cloud. We've also added the capability for customers to buy from the reseller of their choice. And we're finding that customers actually move their workloads to the cloud much faster when they buy from the right ISV at the right time. And that our ISV partners like Glean, they're able to close larger deals faster and more often. But at the same time, being successful at growing with a marketplace is just like many things in business, you get what you put into it, so that's why we started the series and again, thanks Tamar for doing it. To really shine a light on the ISVs who do this well, Glean is one of them, how they've done it, to really help our customers understand why Glean is such a great choice, for customers looking to break down data silos and spread knowledge faster and really to accelerate how they're using AI faster to transform their business. And what you're going to hear I think from Tamar and Glean is that they've done all the necessary three things to really have a successful partnership with us on Marketplace. First, they build unique solutions that customers can only find on GCP. They're one of the most comprehensive search solutions out there to help customers learn from their data and build GenAI applications, which everybody is doing these days very quickly. They market it really well, secondly, to our mutual customers. And they put it in the right place in their product experience with great collateral available for our respective teams, so everybody is educated on how it works. And then we engage in deep co-sell relationships where our field teams work really closely together to make sure we're solving our customer's toughest problems.

>> That's awesome. Tamar, I want to get into how Glean thinks about the marketplace and how your growth has been accelerated by that. But first, let's get some context. TheCUBE research on our side, we see you guys really rising up in the pack there in the marketplace. People are seeing that growth, and congratulations. So explain what do you guys do, because I think this is a great story people should know about. Let's get into what you guys do and then I want drill into some of the marketplace dynamics.

Tamar Yehoshua

>> Thanks so much, John. And Stephen, it is great to be here and to be working with you on our journey. So what does Glean do first? So Glean is a work AI platform, think ChatGPT for your enterprise. We understand all of the context of your enterprise. We look at your Google Workspace documents, Microsoft, AVA 365, Salesforce, Jira, Confluent. Any place where you have siloed data in your SaaS applications, we are able to search across all of them. We have two interfaces, a standard enterprise search where we can find information across all of your documents, and a ChatGPT-like interface where you can ask questions in natural language about information in your enterprise like, "What is the status of this latest project? Who is the salesperson on this account?" and you can get answers in that natural language with generative AI. So it breaks down the information silos and enables you to find information anywhere in your organization which improves productivity and customers are finding it incredibly high ROI to be able to work better and more effectively.

>> That's awesome. And what's the main problem that you guys solve for your customers? What's the big aha moment? What's the secrets and the drivers behind your growth?

Tamar Yehoshua

>> So the first one is finding information. We all are stuck with information in different silos. You're like, "Where did this person send me this? Was it in email? Was it in a document? Was it in a Slack conversation?" So you don't have to remember, you can ask Glean and Glean will find it, but we also have a platform. So in addition to that, we have a platform where you can build either through code or no code. We have APIs to all of the information of your organization. We build essentially a knowledge graph of your organization, so you can use APIs to build productivity apps on top of that, either with code or no code. We also recently introduced Glean apps where you can build with a workflow type builder, no code applications on top of the context of your organization to make yourself even more productive.

>> That's awesome. I want to get into the marketplace. Obviously Google Marketplace, Stephen and I have been talking to their Marvels, their leaders in the space. How have you guys been able to grow your business and accelerate your go-to-market with Google Marketplace?

Tamar Yehoshua

>> It's actually exactly like Stephen said. There are multiple components, one, first build a product that people love and want to use, and then work really closely with Google on two things. The marketing, it's been really effective to co-market together and to sell. So our sellers are promoting marketplace and the marketplace sellers are promoting Glean and we're finding that our customers have an easier experience buying on the marketplace. The procurement is easier, finding us is easier. In fact, from Q1 to Q2, our number of deals on the marketplace doubled. So we're super excited to see that because I think now we're really getting into a groove of understanding how the marketplace can help with us and our customers.

>> What's the-

Stephen Orban

>> If I could just add-

>> Okay. Stephen, go ahead.

Stephen Orban

>> If I could just add to that really quick, John, I think first of all, Glean makes something that's super easy to sell. And as Tamar mentioned, breaking down data silos, you talked to lots of enterprises, I talked to lots of enterprises, it's almost always at the top of their list. "I have all this information, it's spread across all these different systems, I don't know how to make sense of it." So the value proposition is very quick to resonate. And then they integrate with a bunch of our products like BigQuery and Vertex AI and in addition to that, a lot of the other ISVs that we work with. So oftentimes when we're talking to customers, they're taking a collection of things from our marketplace and the fact that Glean can also integrate with Confluent which a customer may have taken to help them break down that silo, makes it super easy. So I think it is hard to overstate just how well integrated, and how much it helps break down the silos and how easy that is to position the customers together.

>> Yeah. And I think-

Tamar Yehoshua

>> For a little bit more background on Glean. Glean started in 2019 as an enterprise search solution, which was using AI, so using AI technologies like Burton, early ML models and vector embeddings for our enterprise search. And it was built fully on GCP, as Stephen said, using BigQuery and using all of the GCP components. Then when ChatGPT came along, GPT-3 came along, we added the ability to ask questions in natural language and to generate answers. And we are then moved to using Vertex AI in addition, and we allow customers to use Gemini 1.3 Pro as well. So we have taken advantage of pretty much most of the GCP stack, and it's worked extremely well for us.

>> Tamar, I love the fact that you guys are just a cutting edge startup. And the joke years ago, Stephen was 10 x engineer. Now you're seeing with AI and startups, you can do a lot more with what's out there at scale. And I think it shouldn't be underestimated. I want to unpack the marketplace's economics and the workflows. I mean, we'll get to the integration in a second. I think that's an even more interesting part. But Tamar, it's not just a directory to host the catalog. It's not like that. Although it is listed. There are real economics here and there's joint marketing, but it's scalable. How do you see this as a modern approach for startups? Because why wouldn't I want to take advantage of joint selling and clean out the credits budget flush or... These are real economics. Talk about how this is a modern scalable solution.

Tamar Yehoshua

>> These are real advantages. If a customer already is allocated money for GCP, and the fact that we can tap into some of their commits is huge, and the fact that the sellers from GCP can work with us to sell Glean alongside what they're selling for G for GCP, this makes it easier for our customers. And that's really what you want at the end of the day, you want a better customer experience.

>> Yes. The integrations become, again, my favorite part is the integrations because that's also unique. First of all, economics. I'm a big fan of the marketplace. Everyone knows I'm pretty biased on that, especially your modern startup. But the key to success is we're in a heterogeneous world. The integrations are critical. Can you guys share the integration aspect of this, the technology, the innovation? What's different? What's unique about it? How does it work?

Tamar Yehoshua

>> Yeah. So what we do with Glean, there's a lot of sensitive data that Glean has in that we ingest all, we have over 100 connectors to your enterprise data, and we want to make sure that our enterprises feel comfortable and so they can host the data in their GCP project. We have a Glean hosted, which is our SaaS offering, all single tenant. We can host it, but we can also host it inside the customer's GCP. And that's incredibly valuable for them. It's more secure. They feel more comfortable with it. So that's a very big selling point for us.

Stephen Orban

>> Yeah. I would just add to that, John. I think the flexibility that Glean offers, whether it to be a managed solution that Glean hosts or something that a customer deploys and manages in their own tenant as Tamar points out, is a big differentiator so much so we actually work together with a very large financial services customer that Google Cloud was already talking to in Asia Pacific, and they had the need to break down their silos and be able to create easily accessible for their knowledge workers. So our team's quickly brought in the Glean team who was able to showcase not only how their solution is awesome, but how it can actually run. This is a regulated entity. How it can run in their own GCP project, which they are now starting to adopt across their business to really give their knowledge workers much more fingertip access to data that would otherwise have taken them hours if not days, to find. So particularly in regulated scenarios, that value proposition resonates quite a bit.

>> I find that the Glean use case very interesting, Tamar, because data is critical and security of that data is awesome. But after you're done training the data, inference and fine tuning become really important aspects of how data evolves going forward. Obviously governance and cataloging of it is really key. This is a key to the security aspect, which is a big part of the enterprise. How does that fit in to your relationship with Google? Is that highlighted by them? Obviously super important, be built from day one. Sensitive data is always that enterprise feature. It's the first conversation.

Tamar Yehoshua

>> Yeah. This is something that Glean has spent five years working on. Making sure that we adhere to the permissions of all of the SaaS apps that we leverage. Making sure that the data is secure and we're a rag-based architecture. So what we do is we read in the data, we adhere to whatever permissions that the original data was in, whether it was in Google or Salesforce or Slack with a private channel, a public channel, what the permissions on the records were. We understand those and we adhere to those. We ingest those into our index, and as I said, the data is stored within a customer's GCP project if that's what they choose. And where RAG-based solution is that we take the query and we match it to the content in order to find the best snippets. And then we match that with generative AI to get an answer in natural language. So we are not fine tuning our own models, we're using more of a RAG-based architecture. But for the search for the retrieval, we use vector embeddings where we do fine tune for a specific organization to make sure that the search is customized and personalized for that organization.

>> So helping us understand the RAG versus the workflow. So using the RAG retrieve augmented generation with Google, using some of their key technologies. But this is the case where the customer brings their data to the table, is that how it works?

Tamar Yehoshua

>> The data is the customer's data. The data is from the SaaS apps that we ingest and that we understand. So it's all the customer's data. We built a hybrid search engine where the hybrid search is we have keyword-based search, and we have semantic-based search to make sure that we understand and we get back the best answers, and we use that through the system. So yes, it's all the customer's data and it's secured in the customer's GCP project.

>> I love this implementation, the best of both worlds. Okay. So Stephen and Tamar, let's get both of you to weigh on this. How does the combined Google Glean solution provide unique value to customers? Connect the dots here because you got a couple things going on both sides here.

Stephen Orban

>> Go ahead. Sorry.

>> Stephen or Tamar, jump ball, whoever wants to go first.

Tamar Yehoshua

>> I was curious how Stephen would say the unique value of Glean, but for us, the unique value is that we really understand the context of the organization. We understand what is going on in the organization because it's not only getting the content, but it's getting you personalized answers. So for example, if I need an onboarding document for somebody in my team, Glean knows who's on my team, who my peers are, what onboarding documents have been used most recently for members of my team, and they won't get an onboarding document that someone, for example from sales was using. They'll know that the onboarding documents of people in my team were the most relevant. So we understand the content. It's personalized to you who's searching for it. It's semantically relevant, and it gets you the answers that you need across all of your information. So our customers just find it the best quality results, plus they don't worry about privacy and permissions because it is fully secure.

Stephen Orban

>> Yeah. I'd say a couple of things just to add to that. So at Google Cloud, we view our role as we're on the customer outcome business. So we want to understand what our customer's hardest technical problems is. Google's mission overall is to solve computer science's hardest problems, and we inherit that in cloud by helping our customers solve their hardest technology challenges. And we talked about some of the key levers here already, breaking down data silos, but I'd also say, it's very timely what Glean does, because every conversation I've been in with the customer for the last at least year and a half, we have over-unders on how long it takes before GenAI comes up as part of that conversation. And it's usually within the first 5 or 10 minutes, what is the GenAI? How are we using it to change our business? And Glean is just a wonderful shortcut to that because you can take all of the data that is spread across in silos in a customer's organization very quickly, connect it up to Glean, so it understands a lot of that content. And then use it, and as Tamar said, a natural language interface and a GenAI interface to not just ask questions of it, to get immediate gratification, but to also start to build applications and start to satisfy a lot of the pressure that these companies are under to continue to differentiate themselves and how are they going to use their own IP in a world where more and more GenAI is going to be used to help companies, including their competitors just move a lot faster. So it's just a super timely, contextually relevant partnership that we're both getting a lot of mileage out of.

>> So Glean is a marketplace customer, got that, but they're also tomorrow used to building on Google's Cloud, with the integrations and offering Vertex Gemini, you're tapping into their capabilities, is that right?

Tamar Yehoshua

>> Yeah. So the way that we work is customer gives a query, and then we understand that query. As I said, we have the RAG-based architecture, we get the snippets and then we use the foundation models to generate the answer. And we work closely with Gemini to make sure that we will validate the latest model. So the customer actually chooses what model they want to use so they can bring their own key. So again, if they have commits already with GCP, they can bring their key to Gemini and whichever model they want. We were using Palm before, we still offer that. Now we offer Gemini. So we do work very closely to make sure that the models will be verified with our stack and so that customers can use those models. And I do also want to build on what Stephen said about customers in that, the biggest thing for us is that the customers are happy and they're getting what they need. And what's been really exciting, Stephen mentioned a bank in Asia, a large financial institution. Also, there's so many customers that are joint customers, they just wanted to take a minute to mention a few. So one that came up through the marketplace that we probably would not have gotten to or maybe it would've taken much longer. Wealthsimple was one, Confluent was another, Reddit was another. These are joint marketplace and Glean customers. There's a much longer list, but just wanted to mention a few.

>> Yeah. So we get this customer, this is important because the customers ultimately vote with their decisions, and that we are at an inflection point. This is a secular trend with generative AI. It's going to change the makeup of all applications and new applications. So what customers do is important. So you said Wealthsimple was one, Reddit, Confluent. What was their key stories? What was the key thing? I'm asking you search finding, what you're looking for. What was the big problem that you solved and what was the impact of their business?

Tamar Yehoshua

>> Finding information. So for example, Wealthsimple, engineers can be much more productive if they can find information when they're, for example, resolving an incident or when you're writing code, the majority of your time spent is actually not writing the code, it's doing the design. It's finding all the previous design documents. It's going through JIRA to understand the bugs that customers had. It's reading all of the documents. So when you can find that information faster and easier, it makes you much more productive. So what we do is we do surveys often for customers before and after. How hard was it for you to find information? And while some, for example, most of the engineers rated very low in that it's really hard to find information. And then rated much higher after Glean, much easier to find information, saving considerable time. And they did estimates of how much time it was saving, the time saved with Glean more than paid for the cost of buying Glean. And that's something that we see in a lot of our customers is that we are saving real time and we have real data that shows the improvement in the productivity for people using Glean.

>> That's the time savings in these markets, reducing the steps that takes to do something, making it easier, understand and intuitive to use is a great business model. I have to ask, Stephen, back in the old days, silos meant databases, big data warehouses. Now silos are Slack chats, Jira, Google Drive. I mean, data's everywhere now. It's like millions of silos. So when you say busing down the silos, Tamar, you mean just unifying this data everywhere? That was a hunt and peck retrieval model in the past. I use hunt and peck, that's way I think about it.

Tamar Yehoshua

>> You're absolutely right. The vast majority, it's something like 90% of data and organizations and it's in unstructured data, which is exactly what you said. It's in Slack or it's in Teams or it's in Google Docs or it's in Jira tickets or Zendesk tickets. So it's unstructured data that is all over the place. Your structured data in your data warehouses, your product analytics, your business metrics, your sales analytics. So we will marry the two and people will be able to query that data as well. But the vast use case and the vast majority of your data actually is in unstructured data.

Stephen Orban

>> If you think about it, John, traditionally breaking down silos, what would that have meant? So if you had 10 systems that you wanted to get data from, you'd have to understand how to get that data from all of those 10 different systems. You'd then have to write some interface to then copy that data somewhere centrally. Now, you've just created a new silo, which is this new centralized place. Now, you have to maintain the 10 connections to all of those 10 different places. Every time Slack changes some interface, that's going to break, you're going to have to disrupt engineers or folks who had did that and Glean just comes and takes that off the table. They're doing that all on the customer's behalf. And it's really breaking down silos without creating a new one, which I think that's maybe been some of the challenges with prior approaches to this problem is breaking down silos by creating a new one. So it's super valuable in terms of a productivity improvement.

>> Well, Tamar, I really like what you do. And again, enterprise search of the yesteryear was one big structured headache, a lot of moving parts, a lot of legacy. Search is the killer app as it proves of the enterprise here. What's changed the most out there? Because it seems that AI is a gift for the old crack the code on enterprise search, but you're not just doing enterprise search now the user experiences are involved. Productivity can be quantified, huge switch. I mean, the game is still the same, enterprise search, okay, bust down the silos, but the apps is the user interface. That's the usability piece.

Tamar Yehoshua

>> 100%. Another thing that's changed is 10, 20 years ago before the Cloud and before SaaS, a lot of your unstructured data was buried on-prem in a system where there were no APIs. So it was very hard to search across all of them. But then when you got to all of the SaaS apps and then everyone has APIs. So all of these SaaS apps have APIs because it's the customer's data. It's not the SaaS apps data. So they enable the customer to access their data through APIs. And this enabled a whole other generation of enterprise search because you could use these APIs. And then take AI even before generative AI, when the BERT model started coming out and vector embeddings, it made it easier to build. You had the tools to build really good enterprise search. So these were some of the changes that happened and then people got used to with ChatGPT querying and querying a natural language and getting an answer. So this change on the enterprise side, enabled people to get comfortable with it.

>> I think you just nailed why you guys are so successful. I think this is a really nexus of a bunch of things coming together. If you think about it's like bring stuff to the front lines from the bowels of these databases. Unearthing the value is awesome. So guys, this is just, again, highlights the market we're in. I have to ask you Tamar, because you know what, I'm always talking to startups out there and you guys are just taking off big time. Things like the marketplace and integrations, what's going on in this modern era is super important to get studied up. They don't teach us in business school. This is not out there. So what's your advice to other entrepreneurs as they look at building good product market fit, tapping things like the marketplace, what are some of the secrets? What are the special knobs you got to turn? I mean, Stephen, always says, "You got to do the 1, 2, 3." I get that, but this is an accelerant. I mean, why wouldn't you? Talk about why it's important to leverage these new modern techniques?

Tamar Yehoshua

>> Well, first of all, advice to startup founders is the same as it's always been. You need to build a product that people need and they love, and you have to be solving a new problem. Some AI startups are taking the technology, "How can I use this to build something cool?" You really need to understand what problem you're solving for customers. That's the most basic thing, and that never changes no matter what the technology stack is. So don't get caught up in the technology stack. Make sure you're focusing on the customer problem. And then once you know you're solving important problem, distribution is key. Distribution is always king. And partnering with a place like the marketplace is so important for distribution. So if you have that product that people find valuable, you need an outlet to make it easy for customers, not only to discover you, but also to buy your software. And that's why the marketplace is so important.

>> I really appreciate your time, Stephen. I'll let you wrap up. Final thoughts. Marketplace what's new? Coming out of the summer into the fall window, a lot of action happening. Market is interesting. It's still ramping up on the AI side, but the pressure on productivity, the pressure on go-to market's huge.

Stephen Orban

>> Yeah. Which is why I think our partnership with Glean is so fast growing. Having said that, it's still early days, and I think what I'm most excited for is that the best is still yet to come. We've built a great foundation and a partnership here with Glean, and we've landed some marquee customers already together. Tamar talked about several of them, but we also have an exciting pipeline and many more customers. We're going to be talking about how they can break down their silos and build their Gen AI applications more quickly. And as I think about how we're also evolving the marketplace and our field incentives and behavior to have more and more conversations with customers about what third party software they're using as we're helping them transform. Adding things like our reseller capabilities, which we brought online earlier this year, to not just help Glean scale through us and our field sales teams, but also all the large resellers like CDW and WWT that are out there. It's just a super exciting time to be in the space, and I think that the best is still yet to come between Glean and us.

>> Awesome. Well, congrats. Tamar, thanks so much for coming on. Stephen, always good to see you again. I love these waves. Every new wave brings new brands, new innovations that move the needle. Thanks so much for sharing and coming on theCUBE. Appreciate it.

Tamar Yehoshua

>> Thanks for having us.

Stephen Orban

>> Thanks for having us, John.

>> Okay. This is theCUBE's presentation of Google Cloud Marketplace Marvel Series. I'm John Furrier, your host of theCUBE here in theCUBE Studios. Thanks for watching.