Databricks Data + AI Summit 2024 | Elliot Shmukler | Anomalo

theCUBE.net

Clips
More from Databricks Data + AI Summit 2024

Elliot Shmukler

CEO

Anomalo

play_circle_outline Introduction to Databricks Data and AI Summit in San Francisco, California

play_circle_outline Importance of unstructured data and its use in generative AI applications

Info

Elliot Shmukler | Anomalo

Elliot Shmukler

CEO Anomalo

search

>> Good afternoon, data community, and welcome back to Databricks Data and AI Summit here in our fabulous hometown of San Francisco, California. My name's Savannah Peterson. Very excited to be joined here with the man himself, John Furrier. John, thanks for bringing me to my first Databricks.

>> Oh, right. This is a great event. It's very technical, which I knew you would love.

>> Yeah, you know I'm a good hardcore geek like that.

>> AI, BI, data quality. This has got everything.

>> It really does have everything.

>> Everything.

>> Including our next fantastic guest. Elliot, thank you so much for joining us today.

>> Thank you so much for having me.

>> Your smile and energy right away when you walked up here, it's contagious. We love that in the afternoon of day two. Feeling the vibe, feeling the vibe. How has the show been for you so far?

>> The show's great. The show's great. Everyone's excited about data and AI, and everyone's excited about data quality, which is what we work on, and it's just great to see everybody. I feel like I'm seeing a lot of old friends, a lot of our customers, a lot of our partners. It's just amazing.

>> Take us through the history. You have history with Databricks, obviously they're an investor in your last round of funding. They've been around, all the high-tech COMMSCI community is involved. Going back to 14 years ago when Big Data came in, Spark Summit 2014 ... '12 I think it was, and then since it's just been one big data community.

>> Yes, it has been. It's amazing. I remember 10 plus years ago, just struggling to use data, just struggling to query data. I was at LinkedIn many years ago.

>> Oh, my gosh. Yes.

>> I couldn't even query more than a day of data at a time.

>> I remember bribing our BI guy at the 3D Printing startup, that I literally would bring him in cookies and stuff to try and get my queries prioritized, and it would still take two weeks to get the data I wanted nine times out of 10.

>> Exactly, exactly. And now in minutes you can have a Databricks lake house and you can query whatever you want. It's just an amazing time.

>> It is a really amazing time. This has obviously got to be a moment for you and the company. Were you anticipating this level of intersection of awesomeness right now, or has the last two years been as thrilling for you as it has been for us?

>> Well, obviously the generative AI revolution, I think surprised everybody. I don't think anyone could have predicted that. We started the company five years ago, and so we definitely thought data was important and more folks were going to be using data and centralizing their data in platforms like Databricks, so we believe that. But the excitement you're seeing now with regards to AI ... Yeah.

>> One of the things, Elliot, we're seeing in the show that we'd love to get your thoughts on, is obviously unstructured data has been a big part of the whole stuff happening. Semi-structured as well has got action. So now to make all that work for generative AI, data quality and governance of the two hot areas. Governance used to be like, "Governance, we'll talk about it." And yawn, but now it's not a yawner. It's actually key. And then obviously data quality has just been chasing your tail. Now it's legit . Take us through the dynamics of now that we have all this unstructured data, where's the action? Where's the work being done? What are you guys doing? Take us through your thoughts on that.

>> Absolutely. So almost all of our customers, and we work for a lot of large organizations, they want to make generative AI work for them. And they're actually using all of the three types of data that you mentioned, John. They're using structured data. You have a customer service chatbot, for example, powered by generative AI. Well, you may want to tell it something about that customer, their name, their last order, their history with the company. That's your structured data, that's data coming out of their data warehouse. Anomalo monitors that data for quality today, but they've also realized that they have these collections of unstructured data. We have a customer, Discover Financial. They have millions of PDFs of credit applications.

>> Oh my gosh, I can't even think about that.

>> Yeah.

>> Yeah.

>> It's crazy, right? We have another customer who's a large insurance company. They have every chat interaction that their customer service agent has ever had with a customer-

>> Holy moly....

>> in their data lake. And so before they were like, "Well, it's nice that we have this data. If we need a particular document, we can find it." But now with generative AI, they can actually analyze that data, and what's more, it's useful if they want build their own model or their own generative AI application, because now you can prompt generative AI applications with data. You can use technologies like RAG retrieval, augmented generation, that's searched through those documents and prompts the model with exactly the right ones. You can fine-tune models. And so they're starting to look at these data sets and saying, "What can we do? What can we do with this data?" And we actually just announced that we have a new product that really helps them get into these unstructured datasets and curate and understand what's in there? Is it high quality, low quality? What should I be using for the future of generative AI applications that I might be thinking about?

>> What about the technology around ... Let me back up.

>> Yeah.

>> One of the things that we were talking about coming into the show was how Hadoop missed the boat a little bit with operations.

>> Yes.

>> It was a hassle. The stack was complex, finding labor and people to run it. What's going on now that addresses that? Because we're hearing simplicity, we're hearing trying to tame the complexity. What's working now and what's different? Why do you see this happening? Are we in a good position to solve some of those problems? Take us through the complexity of the stack.

>> Absolutely, absolutely. So that's actually what's been amazing over the last, let's say 10 years, is the level of automation. I actually remember when I was at LinkedIn, we built out our Hadoop cluster. We had like 150 people on that team building our own cluster. That is ridiculous.

>> But you needed to.

>> Did it work?

>> You needed to.

>> You absolutely had to.

>> Oh, it worked great. It worked great. It was the only way to get key data back. It was amazing. You just don't need that anymore. You can put up an equivalent type of environment with Databricks in minutes. It's all managed for you in that environment. You just do it online. So now the labor-intensive stuff has moved to other areas. For a while, there was a lot of labor-intensive stuff around data quality, establishing rules for your data. A lot of AI and ML has eliminated that. That's a lot of what our product does is we give you a way of monitoring your data and check its data quality without setting up any rules. And the same thing is now happening in generative AI. If you're building an application today, you're doing a lot of manual work. So that's the next area of development, which is when are we going to get frameworks and tools and ways that automate all the annoying manual steps in getting something out the door?

>> I want to automate all the annoying manual steps in everything. I'm curious, we're at their show and it's very cool that you're sitting here with us. What's it like having Databricks as an investor?

>> It's great, obviously-,

>> You have to say that, I guess.

>> They're terrible.

>> It truly is great. It's just great validation for us.

>> That's what I was wondering ... Yeah.

>> When we talk to our Databricks customers, our joint customers, and they know that Databricks has invested in us, it just gives them more confidence. Because you imagine, and this is actually true, that Databricks would not have invested had they not seen our success in their community, seen our growth. They actually even insisted on talking to some customers directly about their use of Anomalo-

>> As well they should....

>> before even considering an investment. It's just an extra bit of validation. We were also very fortunate, earlier this week we won the Databricks Emerging Partner of the Year.

>> Congratulations.

>> Thank you. Very exciting.

>> That's so cool.

>> And so that's just another way for us to very quickly communicate to Databricks customers that we work with Databricks and we work with this community and we're adding value.

>> You guys got a lot of great logos, emerging companies, tech companies that know their stuff. You see a very well-educated data target customer. What's the secret sauce? Why are they working with you? Take us through some of the use cases and some of the success stories.

>> Yeah, absolutely. So the big thing that we bring to our customers is really automation through AI and machine learning. Again, we're a data quality company. We help you monitor your data for quality issues. We help spot those issues, alert you about them. But in the past, what you had to do is actually write rules for your data to detect every single issue. If you didn't write a rule to detect an issue, you couldn't find it.

>> It wasn't detected.

>> It wasn't detected, yeah.

>> It just went on. Yeah.

>> Exactly. And as folks started growing the scale of their data operations, they realized that's just not tenable. So we had one of our big financial services customers estimated that to write all the rules that they would need to monitor their data, they would need 5 million person hours.

>> Holy moly, that is a ton.

>> They actually could probably build a team to do that.

>> That's brute force right there.

>> God, seriously, John. Talk about steamrolling through the situation. That is ... Yeah.

>> So the really distinguishing feature of Anomalo is we can do all of that work for you through machine learning and AI. We actually, instead of asking you to build rules, we have models that are automatically built for your dataset that learn what's normal in your data and are able to spot issues as they arise without you having to tell us where to look or what rule to run.

>> No wonder you're so beloved by your community.

>> We're trying. We're trying.

>> 5 million working hours is a lot of time saved.

>> That's right.

>> That is an incredible time to value ROI on an investment like that, which there's no way any executive wouldn't want to take advantage of that. You've mentioned community a couple times. I'm a big community person. That's one of my angles here on theCUBE. This is obviously a room full of your community, which is awesome. What's the feedback loop like for you in terms of the product roadmap with your community? How frequently are you integrating with them?

>> All the time. We like to keep the cycle as short as possible between ideas and input that we get from our customers and our users and our actual product. And so this is why I actually love coming to these shows. My calendar was actually filled with meetings with folks that are using Anomalo or considering using Anomalo or trialing Anomalo. And so I actually get lots of direct feedback throughout these few days, and that's stuff that I can take back to the team and say, "Everyone was talking about this thing. We need to do that."

>> What are you going to take back to the team this time?

>> Well, everyone's pretty excited about what we're doing with unstructured data.

>> I bet.

>> So that's confirmation that we're going in the right direction. Just a lot of excitement about everyone at this conference is thinking about how do I get ready for AI? How do I get ready for generative AI? And so being able to monitor unstructured data, but also just being able to enforce data quality for all of your data, which is what we do, is pretty critical.

>> Talk about the company, how much you've raised, how big you are, what you guys are trying to do, put the plug in.

>> Yeah, absolutely. So the company's Anomalo. We are a series B company. We've raised about $72 million.

>> Woo.

>> Yeah, some of it from Databricks. And we're about 50 people-

>> That's a strong B....

>> growing very quickly. So we've been consistently doubling our revenue every year, and in May more than doubled our revenue from the prior year.

>> You're crushing it right now.

>> Thank you. Thank you.

>> It sounds a little bit like that hockey stick curve that those VCs love to hear about.

>> Yeah, we're trying.

>> All the investors banging on your door hard right now, trying to get the next round?

>> Some investors are banging on our door. Yeah.

>> It's raining money.

>> Well, I don't know about that, but certainly, hey, trying to take a look. Yeah.

>> Hey, that's exactly how you want it to be.

>> That's right.

>> What's your expression? You can't go out of business if you don't run out of money. Keep the money coming, you keep happy customers.

>> That is what they say. Yeah.

>> Yeah.

>> good. Well, that's cool. Are you guys going to hire any hires? You're probably looking to hire people?

>> Yeah, we're doubling the team this year. So we're going to grow from about 50 people to about 90 or 100 people at the end of the year. So definitely hiring for everything. Whether it's sales or engineering, we have jobs for you.

>> And what a great time to join a company. Series B. Woo-wee.

>> It's pretty early, but not so early that we're going to disappear tomorrow.

>> No, no. You've got the proof of concept rolling and ... Well, far more than that, but-

>> The market fit looks pretty good.

>> Yeah, yeah, yeah.

>> Thank you.

>> What's next? What's next after this event? Next year, what's the vision? What are you trying to do?

>> Fundamentally, we're trying to make it easier to work with data. That's why we got into this. I was telling Savannah earlier that just querying data was so challenging back in the day.

>> You're really bringing back horror memories for me from that era. Yeah, it was a time.

>> It was a time. And now it's easy to query data, but it's still hard to make sure that you're querying the right thing and that you actually get the right result back.

>> Oh, yeah.

>> And it's hard to get it into the right form that you need it in, and sometimes even hard to find the place where you're supposed to query it. And so broadly, we want Anomalo to help you with all of those problems over time, to really be your friend that helps you make sure you're using the right data, you're using it correctly.

>> I love your whole tone. You want to be a friend. You're clearly a very community-focused leader, and it warms my jaded tech heart, which is fantastic. You're a lovely guest, Elliot, so I'm sure we're going to have you back on the show.

>> Oh, thank you.

>> Very curious what happens when we have you on the show this time next year at the next summit? What do you hope to be able to say then that you can't yet say today?

>> I hope to say that we've doubled again on every metric-

>> Or more, maybe....

>> in terms of the team, or more. I hope to have a lot of great generative AI and AI announcements of things that we're doing and how we're helping folks take advantage of this amazing to technology revolution. And I hope to still enjoy the show, because-

>> Yeah, honest statement right there....

>> the event is fun.

>> It is fun. And we look forward to telling that story with you next year, Elliot. Thank you again so much for being here.

>> Absolutely. Thank you for having me.

>> John, always a pleasure. You even got your pocket squares out today. Two in a row.

>> Two in a row.

>> Crushing it, man. You're stepping up the fashion game here on theCUBE. It's a vibe.

>> It's summertime, yeah.

>> Yeah. So your summer class comes out? I like this. I like this. I'm here for that vibe. And I hope all of you are staying classy wherever you're tuning into our two days of coverage here at Databricks Data and AI Summit. My name's Savannah Peterson from San Francisco. You're watching theCUBE, the leading source for enterprise tech news.

Elliot Shmukler | Anomalo

search

>> Good afternoon, data community, and welcome back to Databricks Data and AI Summit here in our fabulous hometown of San Francisco, California. My name's Savannah Peterson. Very excited to be joined here with the man himself, John Furrier. John, thanks for bringing me to my first Databricks.

>> Oh, right. This is a great event. It's very technical, which I knew you would love.

>> Yeah, you know I'm a good hardcore geek like that.

>> AI, BI, data quality. This has got everything.

>> It really does have everything.

>> Everything.

>> Including our next fantastic guest. Elliot, thank you so much for joining us today.

>> Thank you so much for having me.

>> Your smile and energy right away when you walked up here, it's contagious. We love that in the afternoon of day two. Feeling the vibe, feeling the vibe. How has the show been for you so far?

>> The show's great. The show's great. Everyone's excited about data and AI, and everyone's excited about data quality, which is what we work on, and it's just great to see everybody. I feel like I'm seeing a lot of old friends, a lot of our customers, a lot of our partners. It's just amazing.

>> Take us through the history. You have history with Databricks, obviously they're an investor in your last round of funding. They've been around, all the high-tech COMMSCI community is involved. Going back to 14 years ago when Big Data came in, Spark Summit 2014 ... '12 I think it was, and then since it's just been one big data community.

>> Yes, it has been. It's amazing. I remember 10 plus years ago, just struggling to use data, just struggling to query data. I was at LinkedIn many years ago.

>> Oh, my gosh. Yes.

>> I couldn't even query more than a day of data at a time.

>> I remember bribing our BI guy at the 3D Printing startup, that I literally would bring him in cookies and stuff to try and get my queries prioritized, and it would still take two weeks to get the data I wanted nine times out of 10.

>> Exactly, exactly. And now in minutes you can have a Databricks lake house and you can query whatever you want. It's just an amazing time.

>> It is a really amazing time. This has obviously got to be a moment for you and the company. Were you anticipating this level of intersection of awesomeness right now, or has the last two years been as thrilling for you as it has been for us?

>> Well, obviously the generative AI revolution, I think surprised everybody. I don't think anyone could have predicted that. We started the company five years ago, and so we definitely thought data was important and more folks were going to be using data and centralizing their data in platforms like Databricks, so we believe that. But the excitement you're seeing now with regards to AI ... Yeah.

>> One of the things, Elliot, we're seeing in the show that we'd love to get your thoughts on, is obviously unstructured data has been a big part of the whole stuff happening. Semi-structured as well has got action. So now to make all that work for generative AI, data quality and governance of the two hot areas. Governance used to be like, "Governance, we'll talk about it." And yawn, but now it's not a yawner. It's actually key. And then obviously data quality has just been chasing your tail. Now it's legit . Take us through the dynamics of now that we have all this unstructured data, where's the action? Where's the work being done? What are you guys doing? Take us through your thoughts on that.

>> Absolutely. So almost all of our customers, and we work for a lot of large organizations, they want to make generative AI work for them. And they're actually using all of the three types of data that you mentioned, John. They're using structured data. You have a customer service chatbot, for example, powered by generative AI. Well, you may want to tell it something about that customer, their name, their last order, their history with the company. That's your structured data, that's data coming out of their data warehouse. Anomalo monitors that data for quality today, but they've also realized that they have these collections of unstructured data. We have a customer, Discover Financial. They have millions of PDFs of credit applications.

>> Oh my gosh, I can't even think about that.

>> Yeah.

>> Yeah.

>> It's crazy, right? We have another customer who's a large insurance company. They have every chat interaction that their customer service agent has ever had with a customer-

>> Holy moly....

>> in their data lake. And so before they were like, "Well, it's nice that we have this data. If we need a particular document, we can find it." But now with generative AI, they can actually analyze that data, and what's more, it's useful if they want build their own model or their own generative AI application, because now you can prompt generative AI applications with data. You can use technologies like RAG retrieval, augmented generation, that's searched through those documents and prompts the model with exactly the right ones. You can fine-tune models. And so they're starting to look at these data sets and saying, "What can we do? What can we do with this data?" And we actually just announced that we have a new product that really helps them get into these unstructured datasets and curate and understand what's in there? Is it high quality, low quality? What should I be using for the future of generative AI applications that I might be thinking about?

>> What about the technology around ... Let me back up.

>> Yeah.

>> One of the things that we were talking about coming into the show was how Hadoop missed the boat a little bit with operations.

>> Yes.

>> It was a hassle. The stack was complex, finding labor and people to run it. What's going on now that addresses that? Because we're hearing simplicity, we're hearing trying to tame the complexity. What's working now and what's different? Why do you see this happening? Are we in a good position to solve some of those problems? Take us through the complexity of the stack.

>> Absolutely, absolutely. So that's actually what's been amazing over the last, let's say 10 years, is the level of automation. I actually remember when I was at LinkedIn, we built out our Hadoop cluster. We had like 150 people on that team building our own cluster. That is ridiculous.

>> But you needed to.

>> Did it work?

>> You needed to.

>> You absolutely had to.

>> Oh, it worked great. It worked great. It was the only way to get key data back. It was amazing. You just don't need that anymore. You can put up an equivalent type of environment with Databricks in minutes. It's all managed for you in that environment. You just do it online. So now the labor-intensive stuff has moved to other areas. For a while, there was a lot of labor-intensive stuff around data quality, establishing rules for your data. A lot of AI and ML has eliminated that. That's a lot of what our product does is we give you a way of monitoring your data and check its data quality without setting up any rules. And the same thing is now happening in generative AI. If you're building an application today, you're doing a lot of manual work. So that's the next area of development, which is when are we going to get frameworks and tools and ways that automate all the annoying manual steps in getting something out the door?

>> I want to automate all the annoying manual steps in everything. I'm curious, we're at their show and it's very cool that you're sitting here with us. What's it like having Databricks as an investor?

>> It's great, obviously-,

>> You have to say that, I guess.

>> They're terrible.

>> It truly is great. It's just great validation for us.

>> That's what I was wondering ... Yeah.

>> When we talk to our Databricks customers, our joint customers, and they know that Databricks has invested in us, it just gives them more confidence. Because you imagine, and this is actually true, that Databricks would not have invested had they not seen our success in their community, seen our growth. They actually even insisted on talking to some customers directly about their use of Anomalo-

>> As well they should....

>> before even considering an investment. It's just an extra bit of validation. We were also very fortunate, earlier this week we won the Databricks Emerging Partner of the Year.

>> Congratulations.

>> Thank you. Very exciting.

>> That's so cool.

>> And so that's just another way for us to very quickly communicate to Databricks customers that we work with Databricks and we work with this community and we're adding value.

>> You guys got a lot of great logos, emerging companies, tech companies that know their stuff. You see a very well-educated data target customer. What's the secret sauce? Why are they working with you? Take us through some of the use cases and some of the success stories.

>> Yeah, absolutely. So the big thing that we bring to our customers is really automation through AI and machine learning. Again, we're a data quality company. We help you monitor your data for quality issues. We help spot those issues, alert you about them. But in the past, what you had to do is actually write rules for your data to detect every single issue. If you didn't write a rule to detect an issue, you couldn't find it.

>> It wasn't detected.

>> It wasn't detected, yeah.

>> It just went on. Yeah.

>> Exactly. And as folks started growing the scale of their data operations, they realized that's just not tenable. So we had one of our big financial services customers estimated that to write all the rules that they would need to monitor their data, they would need 5 million person hours.

>> Holy moly, that is a ton.

>> They actually could probably build a team to do that.

>> That's brute force right there.

>> God, seriously, John. Talk about steamrolling through the situation. That is ... Yeah.

>> So the really distinguishing feature of Anomalo is we can do all of that work for you through machine learning and AI. We actually, instead of asking you to build rules, we have models that are automatically built for your dataset that learn what's normal in your data and are able to spot issues as they arise without you having to tell us where to look or what rule to run.

>> No wonder you're so beloved by your community.

>> We're trying. We're trying.

>> 5 million working hours is a lot of time saved.

>> That's right.

>> That is an incredible time to value ROI on an investment like that, which there's no way any executive wouldn't want to take advantage of that. You've mentioned community a couple times. I'm a big community person. That's one of my angles here on theCUBE. This is obviously a room full of your community, which is awesome. What's the feedback loop like for you in terms of the product roadmap with your community? How frequently are you integrating with them?

>> All the time. We like to keep the cycle as short as possible between ideas and input that we get from our customers and our users and our actual product. And so this is why I actually love coming to these shows. My calendar was actually filled with meetings with folks that are using Anomalo or considering using Anomalo or trialing Anomalo. And so I actually get lots of direct feedback throughout these few days, and that's stuff that I can take back to the team and say, "Everyone was talking about this thing. We need to do that."

>> What are you going to take back to the team this time?

>> Well, everyone's pretty excited about what we're doing with unstructured data.

>> I bet.

>> So that's confirmation that we're going in the right direction. Just a lot of excitement about everyone at this conference is thinking about how do I get ready for AI? How do I get ready for generative AI? And so being able to monitor unstructured data, but also just being able to enforce data quality for all of your data, which is what we do, is pretty critical.

>> Talk about the company, how much you've raised, how big you are, what you guys are trying to do, put the plug in.

>> Yeah, absolutely. So the company's Anomalo. We are a series B company. We've raised about $72 million.

>> Woo.

>> Yeah, some of it from Databricks. And we're about 50 people-

>> That's a strong B....

>> growing very quickly. So we've been consistently doubling our revenue every year, and in May more than doubled our revenue from the prior year.

>> You're crushing it right now.

>> Thank you. Thank you.

>> It sounds a little bit like that hockey stick curve that those VCs love to hear about.

>> Yeah, we're trying.

>> All the investors banging on your door hard right now, trying to get the next round?

>> Some investors are banging on our door. Yeah.

>> It's raining money.

>> Well, I don't know about that, but certainly, hey, trying to take a look. Yeah.

>> Hey, that's exactly how you want it to be.

>> That's right.

>> What's your expression? You can't go out of business if you don't run out of money. Keep the money coming, you keep happy customers.

>> That is what they say. Yeah.

>> Yeah.

>> good. Well, that's cool. Are you guys going to hire any hires? You're probably looking to hire people?

>> Yeah, we're doubling the team this year. So we're going to grow from about 50 people to about 90 or 100 people at the end of the year. So definitely hiring for everything. Whether it's sales or engineering, we have jobs for you.

>> And what a great time to join a company. Series B. Woo-wee.

>> It's pretty early, but not so early that we're going to disappear tomorrow.

>> No, no. You've got the proof of concept rolling and ... Well, far more than that, but-

>> The market fit looks pretty good.

>> Yeah, yeah, yeah.

>> Thank you.

>> What's next? What's next after this event? Next year, what's the vision? What are you trying to do?

>> Fundamentally, we're trying to make it easier to work with data. That's why we got into this. I was telling Savannah earlier that just querying data was so challenging back in the day.

>> You're really bringing back horror memories for me from that era. Yeah, it was a time.

>> It was a time. And now it's easy to query data, but it's still hard to make sure that you're querying the right thing and that you actually get the right result back.

>> Oh, yeah.

>> And it's hard to get it into the right form that you need it in, and sometimes even hard to find the place where you're supposed to query it. And so broadly, we want Anomalo to help you with all of those problems over time, to really be your friend that helps you make sure you're using the right data, you're using it correctly.

>> I love your whole tone. You want to be a friend. You're clearly a very community-focused leader, and it warms my jaded tech heart, which is fantastic. You're a lovely guest, Elliot, so I'm sure we're going to have you back on the show.

>> Oh, thank you.

>> Very curious what happens when we have you on the show this time next year at the next summit? What do you hope to be able to say then that you can't yet say today?

>> I hope to say that we've doubled again on every metric-

>> Or more, maybe....

>> in terms of the team, or more. I hope to have a lot of great generative AI and AI announcements of things that we're doing and how we're helping folks take advantage of this amazing to technology revolution. And I hope to still enjoy the show, because-

>> Yeah, honest statement right there....

>> the event is fun.

>> It is fun. And we look forward to telling that story with you next year, Elliot. Thank you again so much for being here.

>> Absolutely. Thank you for having me.

>> John, always a pleasure. You even got your pocket squares out today. Two in a row.

>> Two in a row.

>> Crushing it, man. You're stepping up the fashion game here on theCUBE. It's a vibe.

>> It's summertime, yeah.

>> Yeah. So your summer class comes out? I like this. I like this. I'm here for that vibe. And I hope all of you are staying classy wherever you're tuning into our two days of coverage here at Databricks Data and AI Summit. My name's Savannah Peterson from San Francisco. You're watching theCUBE, the leading source for enterprise tech news.