VAST Data Aims to Capture Global Data

VAST Data, a company positioning itself at the forefront of artificial intelligence infrastructure, has set its sights on an ambitious goal: creating technology capable of capturing, contextualizing, and activating the world’s entire data reservoir. Co-founder Jeff Denworth describes this vision as developing a system that can harness global information, process it using statistical or AI-driven tools, and ultimately transform that data into actionable outcomes through AI agents.

During a recent visit to Sydney, Denworth highlighted the company’s collaboration with Australian high-performance computing firm Sharon AI. This partnership involves deploying VAST Data’s Insight Engine and its newly launched AI OS within NEXTDC’s Tier IV M3 data center located in Melbourne. The AI OS, introduced in May 2025, is promoted as a groundbreaking parallel distributed system architecture. It is engineered to fully parallelize AI and analytics tasks, unify separate clusters into a single computing and data cloud, and supply new AI workloads with virtually limitless data from a single, cost-effective storage tier.

Denworth pointed out a significant industry challenge. While businesses are rapidly adopting enterprise AI tools for automation and productivity gains, they often face substantial power demands that traditional data centers were never built to handle. He explained that the joint initiative with Sharon AI is focused on constructing customer pipelines that serve as practical demonstrations of what can be accomplished with modern AI frameworks. These capabilities were recently showcased at an Nvidia conference in Sydney, featuring agents designed to screen resumes for ideal candidates and analyze the effects of climate change on coral reefs.

The company’s Australian client base is both extensive and diverse. Denworth confirmed “dozens” of local customers, mentioning the Walter and Eliza Hall Institute and Dug Technology as examples he could disclose. A broader list of clients reportedly includes prominent names such as Afterpay, the NSW Government, Westpac Group, the Australian Government, and the CSIRO.

Since its establishment in 2016, VAST Data has experienced remarkable expansion. It reached a valuation of $9 billion in a funding round concluded in December 2023. Reports earlier this year indicated the company was pursuing additional investment with a potential valuation soaring to $25 billion. Denworth did not confirm this specific number but noted the business has grown by “almost an order of magnitude” since its last capital raise, justifying a significantly higher worth. He emphasized that the company is in the unusual and strong position of being cash flow positive, growing at an annual rate of three to five times, and holding billions in deferred revenue due to a business model where customers frequently pay in advance for services.

The foundational idea behind VAST Data was to leverage flash storage for AI workloads, enabling simultaneous read and write access to the massive datasets required for AI, a stark contrast to the sequential access limitations of traditional disk storage. Denworth recounted the company’s core realization: for deep learning and AI to become commercially viable, they would fundamentally alter data processing methodologies. This led to the development of a novel storage system designed for massive scalability, low cost, ease of use, and complete parallelism. The architecture was built to process and enrich all data in real-time at any scale, with computation increasingly integrated to allow AI pipelines to run natively on the system.

Industry analyst Gartner recognized VAST Data as a Leader in its October 2024 Magic Quadrant for File and Object Storage, praising its market traction, efficiency at scale, and superior customer experience. The report commended VAST’s strategic partnerships and its use of cost-effective QLC flash and advanced data reduction to deliver high performance for large-scale deployments. However, Gartner also noted some cautions, including the lack of branded integrated appliances, which could deter risk-averse global enterprises, concerns over the frequency of software updates, and the absence of certain enterprise features like synchronous replication and robust hybrid cloud deployment options for production.

In response, Denworth clarified that while VAST initially sold its own appliances, its software is now available on hardware from partners like Cisco, Super Micro, and HPE. He also stated that the limited hybrid cloud traction mentioned by Gartner has since been addressed with effective solutions now available on AWS and Google Cloud. Regarding the rapid software release cycle, Denworth acknowledged it as a “fair comment,” attributing it to the company’s revolutionary technology and its commitment to fast-paced innovation. With a team of roughly 500 developers, VAST Data is pushing the boundaries of modern data processing, aiming to consolidate storage, database, and compute environments into a single, monolithic software stack, an endeavor he claims is unprecedented outside of certain public cloud configurations that often rely on stitched-together open-source components.

(Source: ITWire Australia)