In today’s AI-driven enterprise, speed, context, and scale aren’t just advantages, they are requirements. Traditional architectures that separate compute from data simply can’t keep up with the demands of large language models, agentic AI systems, and real-time inference. That’s why we’re proud to unveil the upcoming WEKA AI Data Platform Appliance—a turnkey solution built on the NVIDIA AI Data Platform design and designed to bring compute closer to data and accelerate the entire AI data pipeline.

What Is the NVIDIA AI Data Platform?

The NVIDIA AI Data Platform is a reference architecture designed for enterprise-scale, combining high-performance storage, accelerated computing, and optimized software to support agentic AI systems. It’s designed to empower AI agents with instant access to enterprise data, enabling faster, more accurate responses across a wide range of workloads.

The NVIDIA AI Data Platform integrates:

The WEKA AI Data Platform Appliance builds on this foundation—delivering a complete, production-ready implementation of the NVIDIA AI Data Platform that’s optimized for deployment across leading server platforms.

Built for the Agentic AI Era

The WEKA AI Data Platform Appliance, now in beta, brings together two of WEKA’s most powerful innovations:

  • WEKA Converged– Run WEKA directly on NVIDIA GPU-accelerated servers to co-locate storage and compute, the converged architecture eliminates data movement delays, maximizes GPU utilization, and drastically simplifies infrastructure—no separate storage tiers, no complex networking, no wasted resources.
  • WARRP (WEKA AI RAG Reference Platform) – A modular, open, scalable architecture designed to streamline the deployment and execution of Retrieval-Augmented Generation (RAG) workloads and optimize enterprise-scale inference.

This isn’t experimental—we already have AI customers running at scale on WEKA’s converged architecture today. From high-performance training clusters to production inference environments, converged is a proven and supported deployment model that drives significant savings. Together, these components align with the NVIDIA software stack to deliver a fully integrated, GPU-ready appliance that’s deployable across your choice of server partners.

A Vision Realized: Storage + Compute Converged

With the WEKA AI Data Platform Appliance, we’re advancing a bold vision for enterprise AI:

  • Converged Compute + Storage: Bring GPUs directly to the data, enabling rapid embedding and minimizing data movement latency.
  • Semantically Searchable by Default: Make enterprise data instantly vectorized and searchable as soon as it lands—no more waiting for post-processing.
  • WEKA is Model Context Protocol (MCP)–ready, an emerging protocol which will enable agentic workloads and swarms of AI agents to communicate directly with the data itself. This transforms storage from a passive repository into an active, queryable, and contextual data layer.
  • Semantic Triggers at Ingest: Enable intelligent workflows triggered by the content of data, not just its arrival.

Designed for Speed, Scale, and Simplicity

The appliance is elastic, scalable, and natively integrates with Kubernetes using the WEKA Kubernetes Operator. This means:

  • Seamless orchestration and lifecycle management
  • Job scheduling for GPU and storage nodes through a unified interface
  • Rapid time to value—especially in disaggregated inference environments

Unlike legacy solutions, this isn’t a patchwork of disconnected components. The WEKA AI Data Platform Appliance is a cohesive, intelligent system that’s easy to deploy and even easier to scale.

Partner-Powered Flexibility

We’ve partnered with leading server vendors to provide you with flexibility and performance options that meet your specific needs. Whether you’re running in a private data center or hybrid cloud, you’ll have the freedom to choose your preferred accelerated computing servers—without compromising on performance or manageability.

The WEKA AI Data Platform Appliance is designed as a best–of–breed, future-proof, elastic architecture that evolves with your AI strategy. As the NVIDIA AI Data Platform continues to expand to incorporate new components like graph databases, advanced vector search, or emerging model frameworks, the WEKA appliance enables you to integrate them seamlessly.

Best of all, there’s no lock-in to any specific component. You retain full flexibility to integrate the best-fit technologies for your stack, now and in the future.

Accelerate Your AI. Simplify Your Stack.

With the WEKA AI Data Platform Appliance, you’re not just buying hardware—you’re unlocking a new operational paradigm. One where storage, compute, and intelligence are integrated by design, not by duct tape. Together, WEKA and NVIDIA are powering the infrastructure behind today’s most advanced AI, enabling agentic systems, accelerating inference, and eliminating data bottlenecks. From RAG to real-time reasoning, we deliver the scale and performance AI demands. Now available in early access.

Deploy RAG at Scale in Minutes—Learn More About WARRP