From Data Storage to Orchestration: VAST unveils AI OS
Data infrastructure company VAST Data has announced the launch of its AI Operating System, a platform designed to power large-scale artificial intelligence workflows, as the firm reported reaching $US2 billion in cumulative bookings faster than any data company in history.
The New York-based company said it achieved nearly five-fold year-over-year growth in the first quarter compared to the same period last year, while maintaining a cash-flow positive business model. The milestone underscores growing enterprise demand for AI infrastructure capable of operating at unprecedented scale.
The VAST AI Operating System represents nearly a decade of development work aimed at creating what the company calls an "intelligent platform architecture" capable of harnessing AI supercomputing resources. Built on VAST's proprietary Disaggregated Shared-Everything (DASE) architecture, the system enables complete parallelization of AI and analytics workloads across distributed computing environments.
"This isn't a product release - it's a milestone in the evolution of computing," said Renen Hallak, VAST Data's founder and CEO.
"We've spent the past decade reimagining how data and intelligence converge. Today, we're proud to unveil the AI Operating System for a world that is no longer built around applications - but around agents."
The platform includes comprehensive distributed system components: a kernel for running platform services across private and public clouds, a runtime for deploying AI agents, realtime event processing infrastructure, messaging systems, and distributed file and database storage for realtime data capture and analytics.
AgentEngine Powers AI Workflows
Central to the new offering is the VAST AgentEngine, an auto-scaling AI agent deployment runtime that provides a low-code environment for building intelligent workflows. The system allows users to select reasoning models, define agent tools, and operationalize AI reasoning processes.
Building on the company's 2024 preview of InsightEngine - a service that extracts context from unstructured data using AI embedding tools - AgentEngine represents the next phase of VAST's AI strategy. While InsightEngine prepares data for AI consumption, AgentEngine enables AI agents to interact with that data in realtime.
The AgentEngine features an AI agent tool server supporting Model Context Protocol (MCP)-compatible tools, allowing agents to invoke data, metadata, functions, web search capabilities, or other agents. The platform enables agents to assume multiple personas with different purposes and security credentials while providing secure, realtime tool access.
The system's scheduler and fault-tolerant queuing mechanisms ensure agent resilience against machine or service failures. VAST also introduced what it calls "massively-scalable agentic workflow observability" through parallel, distributed tracing capabilities that provide developers with unified visibility into complex agentic pipelines.
VAST plans to release a set of open-source agents through AgentEngine, with one new agent launching monthly to accelerate adoption of AI computing. The agents will include both industry-specific personal assistants and general-purpose tools.