Data + AI Summit 2025 | Steve Sobel & Andrew Ferguson, Databricks

Clips
News
More from Data + AI Summit 2025

Steve Sobel

Global Industry Leader-Comms, Media & Entertainment

Databricks

Andrew Ferguson

VP, Databricks Ventures

Databricks

Inside Databricks’ AI revolution: 10 key takeaways from the Data + AI Summit

Artificial intelligence is becoming more intuitive, interactive and embedded, not just in systems, but in the way companies build, think and create. From live coding and autonomous agents to unified governance and automation, Databricks Inc. is moving AI closer to the application layer and aligning it with developer intent.TheCUBE Research’s John Furrier and George Gilbert break down the keynote analysis during Databricks’ Data + AI Summit.“You sort of tell the AI engineer, the software developer, an agent, what you want to do,” said George Gilbert, principal analyst for theCUBE Research. “In the past, all it did was generate code

play_circle_outline Databricks event focused on Data + AI Summit, highlighting the evolution of the platform.

play_circle_outline Investment activities in the startup ecosystem through Databricks Ventures and startup programs discussed.

play_circle_outline Databricks' startup program supports early-stage companies with credits and hands-on assistance.

play_circle_outline Highlighted programs for go-to-market support, co-marketing, and enterprise integration for startups.

Info
Transcript

Steve Sobel & Andrew Ferguson, Databricks

Steve Sobel

Global Industry Leader-Comms, Media & Entertainment Databricks

Andrew Ferguson

VP, Databricks Ventures Databricks

Steve Sobel, global leader, Startup and Venture Programs, at Databricks Inc., and Andrew Ferguson, vice president at Databricks Ventures, join theCUBE’s John Furrier at the Databricks Data + AI Summit 2025. The conversation explores Databricks' growing role in supporting startups and accelerating innovation through its Ventures and ecosystem programs.

Ferguson highlights the evolution of Databricks Ventures and the value of programs such as “Built on Databricks” in providing startups with technical alignment and go-to-market support. Sobel shares how D... Read more

explore Keep Exploring

What notable investment related to startups and venture programs does the text mention? add

What is the Venture program associated with Databricks, and what has been its impact since its inception? add

What type of support and incentives does Databricks offer to organizations that are VC-backed and starting on their platform? add

What programs or initiatives does Databricks have to support companies, particularly startups, in their go-to-market strategies? add

bolt Powered by CUBE AI

Steve Sobel & Andrew Ferguson, Databricks

search

>> Welcome back everyone to the Databricks event. It is the Data + AI Summit. Of course, theCUBE's been covering Databricks since the beginning. It's been quite the journey, but now it's prime time for data out in the open with startups and value creation, and you're seeing the platform evolve with Databricks, obviously enterprise ready. It will become the operating system for AI. Many love what Unity's doing, obviously Iceberg. All the open source, all the open, all the strategies have been working. Steve Sobel's here, Global Leader of Startup and Venture programs. Andrew Ferguson, VP of Databricks Ventures. So they have a venture team.

Andrew Ferguson

>> We do.

>> Andrew, thank you for coming on. Steve, great to see you.

Steve Sobel

>> Of course.

Andrew Ferguson

>> Likewise, it's good to be here.

>> So first of all, I'm really excited about what you guys do, even though it might not be the main headline news, it was a billion dollar investment in San Francisco to make it to AI capital world. But what I really like, what you guys are doing is, you guys have always attracted, kind like the alpha developer in data and infrastructure. Go back to the Databricks history, go back 15 years, Hadoop turns into Spark. Spark turns into Lakehouse. I'm simplifying it, but now you've got Agent Bricks, the apps are hitting. So the progression of the data platform era is now prime time. And Jamie Dimon couldn't get any more evident in the keynote where he's actually saying, if you don't do this now, you'll be left behind. Don't argue, just do it, but do it carefully. So that's going to attract the enterprises. But what's really booming Silicon Valley and in New York is these startups that are thinking differently around data, not having the dogma and the legacy of, well, I got a SaaS app and I got a database and I can talk to the database and I've got to write some software and there's my app, put it in the app store. No, no, no, no, a whole nother ballgame. So I want to ask you guys, as you go in Databricks and you see the keynote, what's the temperature of the market? I mean scope, the size of activity, what's happening? Because AGI house in San Francisco is booming. They've got all these micro communities emerging from Salesforce communities, this community over here, this community over here, Microsoft, Amazon, it's all kind of coming together. Who wants take a stab at that. Andrew, you go first.

Andrew Ferguson

>> I'll take a stab. I mean, just think about the size of the expo floor this year relative to one, two, three years ago. I think that is just a testament to the power of the platform and the amount of interest from everyone from developers, to students, to system integrators, to large-scale companies that want to be a part of our ecosystem. And I think this whole conference is a great visual of that.

>> And how long have you guys been around? Because this year it's really blowing up, but it's not new. The Venture's been around, we've covered it. What, 2020? You guys during COVID.

Andrew Ferguson

>> The Venture program?

>> Yeah.

Andrew Ferguson

>> Yeah, so I started in the middle of 2021. Really recognized that even then four years ago, Databricks is a platform company and having a venture program to invest in ecosystem partners to get them to be better, more innovative and more aligned with Databricks' ecosystem was worth doing. And now three and a half, four years later, we have 40 plus portfolio companies. Everything from Series A to Later that are doing all sorts of interesting things with the Databricks platform technology.

>> You're writing checks.

Andrew Ferguson

>> We are, yeah. We invest alongside amazing financial lead investors like our friends at Andreessen and NEA and so many others. And we write a small check to help them be more successful in our ecosystem.

>> We're going to have Pete Sonsini on from Loud Ventures, which he's partnered with the co-founder Alex-

Andrew Ferguson

>> Yes, absolutely.

>> They have a great fund. The AI stuff's hot. What are you guys seeing right now? Because the progression's accelerating. So you start to see the solo founder, you start to see two or three people, the classic, "Hey, show me your plan." No, you got real demos, you've got real customers coming in. Tell me about the entrepreneurial equation that you guys are enabling it, you're seeing in this new data intelligence wave,

Steve Sobel

>> I think we're seeing the full spectrum. I mean we're seeing one in two person startups that are building out agentic solutions. And by the way, in terms of what we're seeing, agentic is red hot. Everyone is coming up with a different slice of agentic for productivity, for sales, for marketing, call center. I mean you name it, we're seeing a startup that is focusing on this. So we're seeing all the way from the two person startup that's bootstrapped and looking to build an AI application all the way through companies that have received $10, $20 million in seed or series A funding, and everything in between. And certainly we have a startup program. So in addition to Ventures, we have a startup program.

>> And that's what you're running?

Steve Sobel

>> Yep. So we give up to $50,000 of Databricks credit plus free business tier support for organizations that are VC-backed that are beginning their journey on Databricks. We have a few hundred kind of customers that have gone through that program. Really great metrics around, startups are hard. What we found is actually 70% of the startups that end up taking our startup program credits end up having a relationship with Databricks after a year. So really good kind of way to get these early stage companies started on the platform.

>> I love the inbound logistics. You write a check, you get the startup, they don't get the check, they can come in through you. So there's multiple entry points. And what services do you offer? Because we've seen this way before in the cloud where some startups might need some solution architects or someone to come in and help. Is that kind of part of your program? Obviously you're writing... How does that work or is that... Thinking through services.

Steve Sobel

>> So actually one thing that's even more important to call out, we have Ventures, we have our startup program. Today, we also announced our free edition where we're actually investing a hundred million dollars in product usage for educators, for hobbyists, for those that are looking to kind of get their appetite wet on the Databricks platform. So even before they become a startup, we have ways that we can support these early, early stage developers. We're seeing in terms of hands-on support, our general model is very hands-on. So if you're a startup and you're kind coming through the door, we often pair the credits with an account team with solution architects that will help you from the word go, help you set up your workspace, get scales, get use cases.

Andrew Ferguson

>> It's value add, it's high value-add focus.

Steve Sobel

>> Well, it's interesting too, what we've found also with our startup program is the huge request we had from customers was like, we want hands-on support.

>> Got it.

Steve Sobel

>> You guys are an awesome platform, but we need a little bit of help. So we've included free support as part of that, and that was really in response to the feedback we were hearing from our customers.

>> Andrew, I've got to ask you, so what's the criteria now for the inbound evaluation? First of all, I'm assuming there's a lot of inbound.

Andrew Ferguson

>> A lot of inbound, I could fill my calendar with that.

>> So what's the filter?

Andrew Ferguson

>> Yeah, well a lot of the standard filters-

Steve Sobel

>> We need an Andrew Bot.

Andrew Ferguson

>> Yeah, an AI Bot to automate the filtering. I mean a lot of the standard criteria that a financial VC would look at, we look at the quality of the founding team, the market opportunity, the customer traction. We do our own customer diligence to really understand the value that the customer is getting from the company's product. And then if that all checks out, it's really about what is the company's ecosystem fit, that's equally important for us. So are they taking advantage of the Databricks platform in interesting ways? Are they integrating and part of our ISP partner program? Are they executing well on that program? Do they have common customers that we can talk to? So we can do an interesting combination of traditional, external VC diligence and then also asking our customers, our engineers, our field, how the company's doing. And that whole picture comes together to help us form a decision around investment. I

>> I noticed you guys hired Stephen Orban from Google. I interviewed him back when he was at AWS. He's the marketplace master. I'm sure there's a go-to-market opportunity in the mix, or is that more downstream, starts got to get a three feet in a cloud of dust first and then they hit Orban for go-to-market?

Steve Sobel

>> We're super excited to have Stephen come on board. He's an exceptional leader. He's built this literally from the ground up at two of the hyperscalers. I think we're super bullish around where we're going.

>> All kidding aside, we love Stephen, but I mean that's a legit... We're seeing startups really leverage the go-to-market operating leverage with the partnership. So you guys already have enterprise access and these startups right now are having a hard time that we cover. They have enterprise targets, but there's no rallying point around how to fit into a stack. They got Kubernetes nailed down, no problem on the platform engineering, but when they get into an enterprise, what do I rally around to sell to the high SLAs and the complex enterprise? It's a huge... They don't even, they just want move the front and the customers are like, well stay in the POC.

Steve Sobel

>> Yeah.

>> So we've started to see real demand for how do I get into the enterprise? Any thoughts on that or is that more downstream?

Steve Sobel

>> One thing that actually a program that we have out, we have a program called Built on Databricks, and really the intention of the program is how do we give go-to-market support for companies across the lexicon, be it early stage, all the way through enterprise, to do co-marketing, co-sell with them. So we found that particularly for startups that are looking for Databricks to be an active channel, we found that to be a great kind of way in to support these companies as they're looking for channel support and they're growing their business. So that's one thing that we do. We also have the marketplace, that's been a great way to create a way for ISVs to show up within the confines of our product. So I would watch this space.

>> I'm a huge marketplace fan because there's so much leverage.

Steve Sobel

>> Oh, for sure.

>> It really is great. I've got to ask about some of the marketing things. I saw this competition, million dollars. Explain how that works. The announcement's coming up or did it go out already?

Andrew Ferguson

>> It went out on yesterday-

>> Yesterday?

Andrew Ferguson

>> During our start form session-

>> Okay, who's the winner?

Andrew Ferguson

>> The winner is Radiant Craft, which is a healthcare marketing company built on top of Databricks. So super exciting. So Steve and I, this is the second startup challenge that we've put together. The first one was focused on Gen AI use cases on Databricks last fall, and then product credits. And then the second and third place winners also won product credits to help use Databricks more effectively.

>> Yeah, that's awesome. What other cool things you got going on? Can you share what the plans are? Obviously coming out of this event, you're going to have your hands full here. I'm expecting a lot of inbound, there's more, there's a thousand flowers that are blooming out there, so it's going to be hard, I can imagine to filter, but you've got to get them to rise up. What do you guys have to kind of nurture the startup ecosystem? Because it's very fertile right now. A lot of experimentation, a lot of cool things going on.

Steve Sobel

>> A big focus for us is going to be how do we show up and how do we do a better job supporting early stage companies that are looking to build on us, particularly in complex areas where there are high data volumes, think areas like observability. So we're working on what are kind of things that we can do to support that community via various levers and channels that we have. More to come on that through the rest of the year, but certainly something that we're very focused on.

>> Certainly anything you guys need from theCUBE, let us know, we're hitting those startups in New York and Silicon Valley as you know. We see a lot of interest. And it's interesting, I have a theory, I've been saying this on theCUBE pod that there will be a big consumer app coming out of this wave. Right now it's a lot of enterprise focus, but a lot of the startups, I mean a lot of the young guns are under 30, they're targeting the enterprise. And so it's enterprise by default because you've got to have data layer, you've got to have all kinds of infrastructure tuning, you've got to run Kubernetes for orchestration. So they're savvy on this stuff. So it's interesting to see the new market be enterprise is the new cool. I think the consumer thing will be, something will break through, but right now it's hot in the enterprise. You guys seeing the same thing?

Andrew Ferguson

>> Absolutely. I mean, yeah, enterprise used to be boring, but steady Eddy and now it's where the dollars are. I think part of it's, as Ali talked about in the keynote, the transformation potential across all those industry verticals and SaaS application verticals because of AI. And I think there's a huge opportunity for a 10 person team leveraging AI in innovative ways to make a real impact.

Steve Sobel

>> And I think it's to kind of double onto that, I think one thing that's different this year compared to years past, you're seeing, and you can see it on the show floor here, you're seeing so much focus on the application layer.

>> Yeah. That's where the business logic is.

Steve Sobel

>> Yep.

>> That's where the data really kind of has to be targeted domain wise.

Steve Sobel

>> And frankly, that kind of application layer, it's not new for us as a company, but it almost feels like this year with so much of the focus on AI, Agentic, it's almost... Certainly not a coming out party at all for Databricks focusing on enterprise or AI or applications. But I think with announcements like you heard today around Databricks Apps, all of the ecosystem investments we're making, I mean ultimately we're listening to our customers. They want to be able to have Databricks in the hands of their end users, be it internal to their organization or external.

>> I think the vibe coding was really a marker in time that's going to set the tone, because you guys really laid that out well. Okay, the front ends are good, now how do you integrate that in? That's where the scale comes in.

Steve Sobel

>> Totally.

>> So the vibe coding will certainly show the direction for how to code with data and then the rest will kind of fall in. So I'm expecting a ton of new capabilities. Are there any areas that you guys see that are not obvious that people might want to look at, that's kind smoke before the fire, so to speak, before the wave hits? Is there any new areas? I mean, we know the top areas, life science is hot, but commerce is hot, retail's hot. I mean where agents and apps are going to come in-

Steve Sobel

>> I mean if you look at-

>> Or all verticals.

Steve Sobel

>> Our lines of business, I mean financial services has been a rocket ship for us. Healthcare and life sciences is one of our fastest growing verticals. We've always done well. I used to lead our go-to-market for communications and media. It's always been a very strong business for us. So I hate to kind of punt on a specific answer, but we're kind of seeing it across every industry, every segment.

>> Every vertical pretty much is going to be disrupted in a good way.

Andrew Ferguson

>> Or we have had some discussions like in manufacturing, a segment where a lot of old line industries, but where technology, especially if it's easier to deploy easier and faster to develop and adopt, you can have huge efficiency improvements that really can transform some of these companies and make it more efficient.

Steve Sobel

>> Also, one thing I would call out that's been interesting, we're starting to see a lot more interest in B2B applications, going after markets like CRM, or going after markets like CDP, or kind of a fusion of both of them. So we're starting to see, I think a little bit of that looking at the, I don't want to call it a legacy landscape because it's not-

>> Like a modern legacy landscape.

Andrew Ferguson

>> No, it's a data-rich landscape.

Steve Sobel

>> Yes.

>> That's exactly it, yeah.

Andrew Ferguson

>> That's the way I look at it, because you can abstract away the interfaces and also teach an agent to learn the legacy, to be the legacy agent.

Steve Sobel

>> It's fascinating. You're seeing a lot of founders come from some of these traditional larger enterprise companies, that are taking new approaches to something that the legacy enterprise might have done before. And there's many of them. I mean, and there's not a lack of investors that are throwing money at many of them as well.

>> You guys will be our new best friends. We really love what you do. And again, we're focused again, silicone angle, we bet on the enterprise 15 years ago, looked at the world talent, great for us. But the thing that we're focused on is where the value creation is and extraction. So creating value and extracting value will be the flywheel on both sides, customer and side. And if coding and sales and marketing is where the hot areas are right now, because it's low-hanging fruit, if we take that forward as developers developing and then go to market, which is commerce, any customer touch point will be impacted. Be it CRM, ERP, digital twins, physical AI's coming. So I think that those two areas are a tell sign and we're trying to squint through that. So I'm sure you guys will get a lot of data to share.

Steve Sobel

>> It's fast moving. It's literally, the conversations with the startups we were having a year ago. It's like a completely new field of things, signals that you're seeing in the market, at least for the ones that we're talking to.

Andrew Ferguson

>> I actually think that one of the many exciting things about Lakebase and the Neon acquisition is that will open up a whole new set of both startups and use cases on Databricks. So I think we'll get even more visibility into different parts of the tech stack.

>> Yeah, you guys really plugged those acquisitions in nicely and they performed. All right guys, put a quick commercial plug in for the last minute we have left. Start with the ventures. What are you working on? How big is the fund? Are you getting more cash? Does Ollie approve all the deals? What happens?

Andrew Ferguson

>> No, it's a very high bar. So we look at a lot but only invest in a few companies. And as one example, we've got 300 plus tech partners and only 40 investments so far. But very full pipeline, partially because we've got a great brand and a great ecosystem to work with.

>> You're intentional on your deals. It's-

Andrew Ferguson

>> Very intentional.

>> It's not like everything's coming through you're really... All right, Steve, give a plug for what you're working on.

Steve Sobel

>> Startup program. We offer up to $50,000 in Databricks credits for VC-backed startups, plus free business tier support. We have a few hundred customers that have gone through this program successfully graduated and are building their applications on Databricks, are using any of our products. So great way for startups to get started on our platform and look forward to working with many more startups in the years to come.

>> Steve, thanks so much, Andrew, thanks so much.

Andrew Ferguson

>> Thank you.

Steve Sobel

>> Thanks John.

>> Databricks Ventures and the Startup program, it's a Cambrian explosion of applications and new ventures to be forming, and transforming existing business models. So stay tuned, follow theCUBE. We'll be right back for more live coverage here at Data + AI after this short break.