Vast Data lands $118M to grow its data storage platform for AI workloads

Vast Data, to make an obvious pun, is raising vast sums of cash.

The New York-based startup, which provides a scale-out, unstructured data storage solution designed to eliminate tiered storage (i.e. setups that move data between high- and low-cost storage hardware), today announced that it secured $118 million in a Series E round led by Fidelity Ventures with participation from New Enterprise Associates, BOND Capital, Drive Capital, Nvidia, Dell, Goldman Sachs, Tiger Global, Commonfund, Norwest, 83North, Greenfield and Next47.

The round values Vast at $9.1 billion post-money, and brings the startup’s total raised to $381 million.

“The explosion of interest in AI and the need for modern infrastructure that can support these workloads in the last year has been a boon for Vast’s business and positions the company for continued growth and adoption with the enterprise,” Vast co-founder and CEO Renen Hallak told TechCrunch in an email interview. “Given the future-proof nature of Vast’s offering, data-driven organizations see Vast as a valuable investment in the future of their business.”

Hallak co-founded Vast in 2016 with Jeff Denworth, Shachar Fienblit (who previously held leadership roles at Kaminario and IBM) and Alon Horev (formerly at Cisco and IBM). The way Hallak tells it, the co-founders shared a vision of creating a next-gen data management platform — one that leveraged commodity hardware to deliver faster access to bigger data sets for AI workloads.

Vast’s founding team subsequently designed a new storage architecture and software infrastructure layer, operating in stealth until 2019, when the company began selling to customers.

Today, Vast unifies storage, database and compute engine services in a platform built to power AI and GPU-accelerated workloads across datacenters and clouds. Customers can use Vast to manage unstructured and structured data across their preferred private, public or hybrid clouds — data ranging from videos and images to text, data streams and edge device data.

“Stitching together legacy enterprise infrastructure is time-consuming and complex, and its inefficiencies make it an expensive endeavor,” Hallak said. “The legacy cloud recipe for building AI infrastructure comprises disparate technologies that, due to their underlying architecture, don’t take full advantage of modern technologies that offer improved performance, simplified operations and cost-savings … [And] without the right infrastructure in place, organizations can’t efficiently enable their AI and GPU-powered investments with the data access needed for AI and deep learning.”

While Vast has competition in vendors like Databricks, Hallak asserts that it has substantial first-mover advantage. There’s some truth to that it seems, judging by Vast’s books.

Vast’s annual recurring revenue now stands at $200 million and the company, which recently inked a strategic partnership with HPE, is growing 3.3x year over year, Hallak says. Cash flow has been positive for the last 12 months while Vast’s customer base has grown to include brands like Pixar and Zoom.

Now with more than 700 employees worldwide, Vast plans to put the new tranche toward expanding its business reach with an emphasis on Asia Pacific, the Middle East and Europe.

“Vast is a software company that operates on commodity hardware, so the pandemic and supply chain issues that plagued many businesses in the last few years did not have a material impact on Vast or its partners and customers,” Hallak said. “While Vast has continued to grow, scale and operate efficiently, this new investment will further advance Vast’s mission to deliver a new category of infrastructure that puts data at the center of how systems think, react and discover.”