Vast Data Founder On New $118M Funding And Why It’s Going To ‘Sit In The Bank’

‘By working with the largest customers, we don’t need to build the same size sales and marketing team as you typically find for startups who come into the market selling to the low end of the market. And because of this, we’ve managed to build a business that doesn’t burn venture capital. Actually, we haven’t spent our series B, series C, or series D. And now, we don’t anticipate spending anything from series E. That leaves us with over $300 million in the bank,’ says Vast Data Co-founder Jeff Denworth.

New $118 Million Funding: Nice To Have, Not Need To Have

This week was a big one for one of the tech industry’s largest “unicorn,” Vast Data. The New York-based developer of a platform which combines storage, database, and compute to power AI and GPU-accelerated applications, unveiled its series E round of funding that not only brought it an extra $118 million in cash, but also nearly tripled its valuation to $9.1 billion.

With the funding round, Vast Data now has $300 million in the bank. But as a company that is already cashflow positive, all that money is just sitting around collecting interest, said Jeff Denworth, a co-founder of the company and the executive overseeing its marketing and product management teams.

For Vast Data, which has yet to spend all the money from its last three funding rounds, the new E round of funding is less about filling the coffers and more about showing the company’s maturity, Denworth told CRN.

Given the small size of the new funding round only amounts to about 1 percent of Vast Data’s total valuation, it was an easy way to significantly increase that valuation while raising the visibility of the company in the market, he said.

“These types of events are very important because there are very few companies that are actually doing what are called ‘up rounds’ in the market and seeing material increases in valuation,” he said. “And so it’s essentially our yardstick that we plant in the market at any given time to say, ‘Look at what’s happening. We’re being validated by some of the best private market investors.’ We’re not a public company yet. And the best way we have to express our business performance is to show the world through these valuating events.”

Not that Vast Data really needed that much additional external validation. The company works closely with Nvidia, which also happens to be a Vast Data investor, to bring to market a platform designed to work with AI platforms around large language models that can use unstructured data.

“There’s like 20 times more unstructured data in the world than there is data sitting in databases and data warehouses, and which has largely been untouched or unusable by most organizations,” Denworth said. “You can’t go back to your files and ask new questions that make sense because the tools weren’t there. Now, with GPUs and neural networks, you have these technologies that are capable of understanding data that comes from the natural world. And so AI is a big trigger.”

Here’s more of CRN’s conversation with Denworth.

How do you define Vast Data?

Vast is building a data platform for the era of AI. A lot of people use the term ‘data platform’ in the marketplace. We define it as a unified computational environment, database environment, storage environment, all rolled into software that can be deployed on premises, on commodity infrastructure, and in the cloud. And we stitch all the different places that you may want to compute together. The really unique thing about us as compared to some of the other data platforms in the market that you may know, products like Snowflake or Databricks, is that we are thematically aligned on this idea that you can integrate everything to make it really easy for customers to compute. Our product is really designed to be optimized for unstructured data, or data like files and objects that you can think of as videos, imagery, sound, things that don’t naturally conform to an enterprise data warehouse or database.

If you think about the unstructured data market opportunity, there’s like 20 times more unstructured data in the world than there is data sitting in databases and data warehouses, and which has largely been untouched or unusable by most organizations. You can’t go back to your files and ask new questions that make sense because the tools weren’t there. Now, with GPUs and neural networks, you have these technologies that are capable of understanding data that comes from the natural world. And so AI is a big trigger. And we sit at the center of this whole event in the marketplace where customers are looking to now organize this data, serve this data up to GPUs, process on this data using new tools. And that’s what we designed the Vast Data platform to do.

We’ve been selling for many years, working with a lot of people that were doing AI. And up to the point the ChatGPT explosion happened, it was a very good growth business. But now that some of the largest AI clouds and some of the largest AI-driven enterprises are deployed on our system, we’ve been seeing an accelerated growth rate on top of an already big and mature business. And so that gets a lot of investors really excited.

How is what Vast Data does different from hyperconverged infrastructure?

There’s a parallel there. But the difference is that hyperconverged infrastructure focuses on things like virtualization management more than it does data management. Our system offers the tight coupling of a very high performance file and object storage system with a transactional and analytical database. Imagine you have a deep learning application doing things like looking at photos for cats. You have a file in the form of an image that may have a cat in it, but the only way that you know how to ask questions of your datasets is when those labels and those learnings are annotated and cataloged in a large-scale, high-performance database. And so when these two things come together, you not only have the raw data payload, but you also have the contextual layer in that database working together. And that’s where things get really interesting. Unstructured data for 20 years has never been able to be put in a data warehouse or things like that. But now that we can actually go curate and annotate and understand this data. The information has to live somewhere. And that’s where the Vast database comes in.

You mentioned that Vast Data’s technology is GPU-based. Whose GPUs do you use?

We’re a software vendor. We don’t actually ship customers hardware. But Nvidia is an investor. We’ve been working with them for many years. We today power some of the world’s largest Nvidia computers. Considering they’re the market leader, a lot of the work that we do is with them.

What other strategic investors does Vast Data have?

So we have two others, Dell Technologies and Goldman Sachs.

You also said customers work with a lot of tools that work with Vast Data’s technology to get information from the unstructured data that’s stored on Vast. What are some of those tools?

If you look out in the market, the people that are deploying AI in anger, there’s some massive investments that are being made right now, like single customers allocating trillions of dollars of capital to address this. And so we view it as a stack that consists of, at the computational layer, Nvidia. And we saw a collection of real companies, and when I say ‘real,’ like valuations of $5 billion, that are seeing material increases in their valuation throughout the year. So Nvidia is obviously on a run. But then if you look at the platforms that deploy these systems, there’s a [cloud-based GPU] company called CoreWeave that just a few days ago announced they have materially grown their valuation. Turns out they’re a partner of ours.

So you have the cloud platform. You have the compute platform. Then you have the data layer. We’re working with CoreWeave, but we’re also working with some of these other new AI clouds that are starting to creep into the market like Core42, based in the Middle East, now moving into the United States and making massive investments. Lambda Labs, one of the other big emerging AI developer cloud leaders. So now you’ve got the cloud layer.

And the final piece is where the applications are being deployed on top of this stuff. Here, you largely have generative AI large language models, but I can’t disclose any partners. We can talk about a few through the work we do with CoreWeave. I think the way you should think about it is, of the biggest LLM companies that we see in the marketplace that are also incrementing value, there are only two that have announced that they’ve more than doubled valuation over $5 billion dollars or greater. They’re OpenAI and Anthropic.

So this is a very select class of organizations. And if you stack them all up, and we see this as the new AI stack, you’ve got the applications being the LLMs, you’ve got the computational layer being Nvidia, you’ve got the cloud layer being companies like CoreWeave, and then the data layer that cuts across all these different clouds it’s being deployed on in the enterprise is Vast Data.

Vast Data just closed its series E round of funding for $118 million, giving it total funding of $381 million and a valuation of $9.1 billion. What are you going to do with this new funding?

Nothing. It sits in the bank and collects interest. These fundraising rounds are also used to generate interest in Vast in the market. We built this really unique business where we specifically cater to the world’s largest data consumers, and so our average selling price is quite high, over $1 million dollars. And by working with the largest customers, we don’t need to build the same size sales and marketing team as you typically find for startups who come into the market selling to the low end of the market. And because of this, we’ve managed to build a business that doesn’t burn venture capital. Actually, we haven’t spent our series B, series C, or series D. And now, we don’t anticipate spending anything from series E. That leaves us with over $300 million in the bank.

And the question is, why did we do it? Well, if you do the math, $100 million dollars on top of $9 billion means that for just a little bit more than 1 percent dilution, we essentially nearly tripled our valuation. And from a market visibility perspective, these types of events are very important because there are very few companies that are actually doing what are called ‘up rounds’ in the market and seeing material increases in valuation. And so it’s essentially our yardstick that we plant in the market at any given time to say, ‘Look at what’s happening. We’re being validated by some of the best private market investors.’ We’re not a public company yet. And the best way we have to express our business performance is to show the world through these valuating events. So it’s just going to sit in the bank.

There’s talk around Vast Data doing an IPO. Any time table?

What I can tell you is that we’re building a very professional organization. We’re now more than 700 people around the world. We try to run the business like a public company. We close the quarters, and we’re putting in the accounting controls so that we can go public when we choose to do so. And so now it’s just a matter of time. We’ll make the right moves when the time is right.

Vast Data is cashflow positive, right?

Yep. That’s the byproduct of our business model: high average selling price, customers that buy repeatedly. We just don’t have the same expenses that companies are selling to the low end of the market do. We don’t have a ton of sales people. We don’t have a ton of marketing people. The team is relatively lean, given the amount of revenue that we’re creating.

Is the company actually profitable?

Not profitable yet. But we’re cashflow positive.

How does Vast Data work with indirect channels?

We are a channel-first company. We have close to 500 engaged partners around the world. We do all of our business through channel partners and service providers. And there’s a new element to the story that’s starting to emerge in the last six months: systems integrators, which historically we haven’t been working with too much. We’ll be making some big announcements about some large systems integrator collaborations over the next I would say two quarters. So you’re going to see some big news from us here. You’ll also see an amplification of the partnership that we’ve established with [St. Louis-based] WWT (World Wide Technology), which obviously is like the gold standard.

We’re working hard to essentially engage partners like this that are realizing that AI and deep learning are becoming a radically disruptive market event, the tool sets need to change, and the platforms need to change. And we’re helping every enterprise modernize their business and their data so everything can be ready for AI when they’re ready.