Production-ready AI pipelines arrived this week as VAST Data unveiled Foundation Stacks, an open-source library designed to industrialize NVIDIA's AI Blueprints for enterprise deployment. Announced at NVIDIA's GTC 2026 conference in San Jose, the new framework extends NVIDIA's reference architectures into repeatable implementations that run natively on VAST's AI Operating System.
The launch addresses a critical bottleneck in enterprise AI adoption. While NVIDIA AI Blueprints provide starting points for building advanced applications, organizations still face extensive integration work before these systems can run securely in production.
Teams typically must stitch together fragmented infrastructure, orchestration layers, and data services to make AI applications production-ready.
VAST Foundation Stacks solve this by delivering pre-built templates that unify data access, database services, compute orchestration, eventing, and pipeline execution in a single environment. These stacks can be deployed anywhere the VAST AI OS runs, in the cloud or on-premises via VAST's newly announced CNode-X platforms, as part of the NVIDIA AI Data Platform reference design.
The first two Foundation Stacks target specific enterprise use cases. The Video Search and Summarization stack enables organizations to ingest massive volumes of live or archived video and extract insights through semantic indexing and interactive Q&A. The AI-Q based stack provides a foundation for building custom AI researchers that operate across private enterprise data sources, synthesizing hours of research in minutes while maintaining persistent and secure context.
"NVIDIA AI Blueprints have given the market an important starting point for building next-generation applications, but enterprises still need a production-ready way to deploy and operate those capabilities at scale," said John Mao, Vice President of Global Technology Alliances at VAST Data.
The company is taking architectural patterns behind leading NVIDIA Blueprints and giving customers a faster path from experimentation to production for scalable AI pipelines and agentic systems.
The announcement comes amid NVIDIA CEO Jensen Huang's projection of at least $1 trillion in revenue from 2025 through 2027 during his GTC keynote. Huang emphasized that "every single company in the world today has to have an OpenClaw strategy," referring to what he called "the most popular open source project in the history of humanity" for agentic computing.
VAST plans to release additional Foundation Stacks in coming months, including industry-focused examples. The initial implementations will be available through a public GitHub repository with interactive demos and planned sandbox environments for customers and partners as part of NVIDIA's projected $1 trillion revenue ecosystem from 2025 through 2027.
