Vast Data has closed its new Series F round of funding to the tune of $1 billion, giving the company a $30 billion valuation—a massive foundation on which to increase its presence in the AI industry with what it calls its AI operating system, which moves beyond its high-performance storage technology to include integrating databases, compute orchestration, and a suite of tools tailored for agentic workloads.
Vast Data may once have been known as a storage company but now sees itself as an AI operating system provider.
And when a company like Vast Data, which Tuesday unveiled the close of a new round of funding that brought it $1 billion in cash and a massive valuation of $30 billion, it is free to define itself however it wishes.
However, according to CEO Renen Hallak, that definition as an AI operating system is exactly what Vast Data has become.
[Related: NetApp Sues Former CTO, Alleges He Took Trade Secrets To Rival Vast Data]
“We’re the AI operating system company,” he told CRN. “There’s a new stack that’s getting built. Jensen [Huang, Nvidia founder and CEO,] defined it perfectly. It’s power, chips, software infrastructure, models, and applications. Those are the five layers. We are building that middle layer.”
Seven years ago, Vast Data emerged with storage technology that aimed to provide lightning-fast access to vast amounts of information, a move that laid the foundation for the company becoming a cornerstone for early AI research and training, Hallak said. And as AI evolved, so did the demands on data platforms. Over time, clients’ needs pushed Vast Data to expand its offerings, integrating databases, compute orchestration, and a suite of tools tailored for agentic workloads.
The company’s commitment to innovation is evident in its approach to technology development, Hallak said.
“It was all in-house,” he said. “We can’t take any open source or acquire companies and bolt on their technologies because every one of our pieces is based on our disaggregated, shared-everything architecture. That’s how we get to the level of scale and performance and resilience and simplicity required in this new era.”
There’s a lot going on at Vast Data, the AI market, and the massive investments in AI focused companies. To learn more, read CRN’s complete conversation with Hallak, which was lightly edited for clarity.
How do you define Vast Data?
We’re the AI operating system company. There’s a new stack that’s getting built. Jensen [Huang, Nvidia founder and CEO,] defined it perfectly. It’s power, chips, software infrastructure, models, and applications. Those are the five layers. We are building that middle layer. We started, of course, from storage and databases and building a data platform, but today we’re also adding abilities around compute orchestration and security and providing all the tools that are required for agentic workloads to be abstracted away from the underlying hardware as part of this new stack.
Before we get to the news, you defined Vast Data as an AI operating company, not a storage company or an AI storage company. Talk a little bit about that because when most people think of Vast Data, they think of Vast Data the storage company.
That’s where we started. The first thing we wanted to solve for seven years ago, when we came out with our first product, was to give very fast access to very large amounts of data. We felt that was the anchor of the data center. We felt that was the crux of what needed to be done to enable these new AI training workloads. And over time, as AI moved from training to inference to fine tuning and reinforced learning, as we moved from AI labs into enterprises and clouds, we realized that you need a lot more than just new unstructured data in order to make AI work. And so our customers, who are always on the bleeding edge of AI, kept asking us for more capabilities. They asked us for a new type of database to take advantage of the architecture that we built, and a new way of scheduling different types of jobs across different types of processors across geographies. And so we added the database. We added the DataEngine. We added the data space. We added the InsightEngine and the AgentEngine. And all of these parts fill in the software infrastructure stack that is required for this new era. And that’s why we shifted from storage to data platform to now AI operating systems. We feel that it’s a better characterization of what we do today.
These different technologies that you’ve added over time, were they developed in-house or via acquisitions?
It was all in-house. We can’t take any open source or acquire companies and bolt on their technologies because every one of our pieces is based on our disaggregated, shared-everything architecture. That’s how we get to the level of scale and performance and resilience and simplicity required in this new era. And so for every one of these things that we added, we had to build it from scratch based on our architecture, and it’s all homegrown.
Given all this, you’ve got some news, a billion-dollar funding round that is already completely funded.
Yeah, we’ve not done any rounds for the last couple of years. The last one that we did was a round E at a $9-billion valuation, and now we’ve completed our round F at a $30 billion valuation. Our company is growing fast, and this is another indication of that.
Who invested in this round?
We have a bunch of new investors, like [global investment firm] Access Industries. We have a bunch of existing investors like Fidelity and Nvidia, and the drive is being led by a fund called Drive Capital. It’s the biggest round we’ve ever done, both in terms of dollars and in terms of number of new investors that are joining. I think as we move forward we’re seeing a lot more interest in us, especially as we become the de facto standard for this important part of the new AI stack.
This is not only the biggest investment round for Vast Data. It’s one of the biggest investment rounds ever. It’s coming at the same time that an AI company like OpenAI is looking at an IPO with a valuation of up to $1 trillion. There’s a lot of money going into AI now.
There is. I think everybody’s realizing that AI is going to take over a lot of industries. It’s starting with software, and it’ll continue across the knowledge work industries. And with physical AI, we’re going to see more robots take over industries that are more in the physical universe. And all of that requires a lot of compute. It requires a lot of data. It requires new models. You mentioned OpenAI, which is the frontier model builder. And it requires new applications that need to be written.
Chatbots are the first applications that we see, but there’s going to be a lot more applications getting built on top of these models and on top of this new infrastructure. And to enable those new applications, we need an operating system. We need to make it easy for them. We need to make it secure. We need to make it widespread in the sense that every enterprise can start using AI in a way that would help improve their businesses.
You mentioned Nvidia as one of the investors in this round. Are there any other strategic investors in Vast Data?
In fact, there are a few strategic investors that are joining this round, but I’m not sure they’re allowing us to use their names. I don’t want to state names that I’m not allowed to say.
So the names of all investors in this round have not been released?
Not all of them. Some of them told us that it’s OK, and we’re listing those. As others are confirmed, we’ll make them public.
What does that $30 billion valuation mean to Vast Data and to the AI industry in general?
It’s just an indication of how fast we’ve grown. We are today attached to millions of GPUs. Wherever these large AI factories get built, we tend to be the system that’s powering them, whether it’s in an AI cloud or in an AI lab. More recently, a lot of enterprises have started to use AI in production and have been adopting our system. In terms of just sheer amount of data, we’ve been growing. We’ve been tripling year over year consistently, and that brings us to a very large part of the market. And I think this new valuation is an indication of all of those things.
Can you share any financial numbers with us?
As part of this announcement, we’re sharing that we’ve surpassed $4 billion in sales total over the years. We do multi-year deals, and sell three-year licenses or five-year licenses, and so our annual recurring revenue is smaller as a result. But we’re also sharing that we exited last year with more than half a billion dollars in annual recurring revenue. But yeah, those are two indications of very fast growth.
Is Vast Data profitable, either in terms of net profit or cash flow positive?
We are both. We’ve been cash flow positive for a very long time, pretty much since we started selling, and we are profitable from a GAAP perspective as well.
When can we expect to hear about an IPO?
That’s a good question. We don’t have a date. We haven’t actually started an IPO process. We haven’t gone out and hired bankers or anything of that sort just yet. What we have been doing is preparing ourselves, preparing our systems, preparing our company, such that when we do decide to go public, we can continue focusing on innovation and our customers, which is what we love to do. We love building great products, and we love putting those in the hands of our customers who I consider our partners. Amy Shapero, our first CFO, joined a couple of years ago from Shopify, and she’s been getting us ready. My expectation is that we will be ready later this year, and then we get to decide if and when we want to pull that trigger.
Are Vast Data’s sales primarily direct or indirect?
It depends on how you consider direct and indirect. For example, we sell to a lot of neoclouds. Their end users end up using the system. So I consider them a channel. If you consider a neocloud channel, then a lot of our sales are indirect through that channel. HPE OEMs our solution. Cisco has OEMed our solution. Those are other pathways into the consumer. Even when we sell direct to an end user, more than 95 percent of the cases are through value-added resellers. So within that framework, nearly all of our sales are indirect, although we use various channels to take us to market.
We still have a very capable sales force helping all of those channels make the sale and get to the end user and convince them that our technology is the one they should use, but we lean heavily on the channel, as you can tell.
What’s the typical process a customer who decides to adopt Vast Data technology goes through once the ink has dried on the contract?
Usually they start by moving data over, moving unstructured and structured data into our database on day one. It doesn’t take a long time to move data because we support very standard interfaces. On day one, they get a system that is more scalable, faster, easier to use, and more cost-effective than what they had before. Over time, what they realize is that they can now do more with their information, and they start to move more and more workloads onto us. Those workloads become faster, become easier, so that all of that information becomes accessible to AI workloads. And once they start adopting AI workloads, that’s when they start leveraging our DataEngine and InsightEngine for RAG and AgentEngine that allows them to have observability and reproducibility into what their agents are doing. It’s a gradual process. Usually, we will land with one workload with one data set and then expand into a lot of other workloads and data sets and enable this bridge into the world of AI.
Earlier we talked about acquisitions, and you said that your company focuses more on developing your technologies internally. Have you ever made any acquisitions? I’m not aware of any.
Not in terms of acquiring IP. We did make one acqui-hire of a small group out of Iceland that is now building part of our system with the larger R&D team, but that was just a small one.
Could you see the company making acquisitions in the future? And if so, what would you be looking for?
That’s a really interesting question. I think for an acquisition to make sense, a lot of stars need to align. The first one is cultural. We want a company that fits our values and principles and operates in a similar way such that there will be a good cultural fit. The second is technology, and as I said, so far, it’s been hard for us to find outside technology that we can just bolt into our system because our architecture is so different. But perhaps, as we veer into areas of security for AI that are a little bit higher up the stack, maybe something like that could make sense. And then, of course, the third one needs to be price. And so far, we haven’t found a company that fits all of those criteria. But as we grow and as we generate a lot more cash, that could make sense in the future.
Now I can see all the AI security companies’ bosses calling their marketing people and asking, ‘What do we need to do to get Renen Hallak’s attention?’ So last question, what are your strategic priorities for the rest of 2026?
We want to keep growing. We’ve gotten used to tripling every year. We hope to at least triple again this year. The demand is there. Everybody is realizing that AI requires this new architecture that we’ve built, and so we’re getting pulled into a lot of very interesting projects, and our customer base is growing fast and growing with us very quickly. We started by focusing on the most data-intensive organizations on the planet a few years back, and as we progress, we want to expand into other parts of the market. We do that through partnerships. We’re now partnering with the big clouds, the hyperscalers, for them to offer our services and who are now growing very quickly in the enterprise as enterprises adopt AI into production. And so those are some of the things that we’re focused on. At the same time, we’re growing geographically. We’re adding new countries that we weren’t in before. We’re growing up the stack in the sense that we’re adding more services to our platform that we didn’t have before. And all of those are growth factors and trajectories that we hope will materialize in 2026.
