Providing the data foundations to power the future of AI applications

The AI world is accelerating, posing a huge challenge for engineering and data platform teams: how to build a responsive, scalable, and cost-effective data infrastructure for AI. Often, adaptability comes with high costs and operational burdens.

At NetApp, we have a better way. The NetApp INSIGHT^® 2025 announcements reveal our unified vision: to provide the data foundations that empower customers like to you build powerful AI applications without the high costs or vendor lock-in. With NetApp^® solutions, you can use open-source to store, to stream, to analyze, to search, and to orchestrate your data with superior flexibility—on premises or in any cloud.

We're excited to show you how our latest innovations can lower your operational overhead, scale up into AI production deployments, and give you complete control over your data ecosystem.

Modern AI applications, especially those that use large language models (LLMs), require a flexible approach to data management. The need to ground these models with specific, up-to-date domain knowledge means that your data platform must be more than just a passive repository. It must be an active, integrated component of your AI pipeline that can efficiently and securely handle complex search queries, real-time data streams, and massive datasets.

In trying to achieve that goal, many organizations find themselves wrestling with proprietary solutions that create vendor lock-in, driving up costs and limiting performance. As your AI strategy evolves, you need a data platform that can evolve with you—not one that holds you back. Meeting this need is where the power of open source truly shines, offering a path to innovation without compromise.

"Your data platform can be your greatest accelerator or your biggest bottleneck,” says Ben Slater, VP and General Manager of NetApp Instaclustr^®. “We find that leaders are most concerned with maintaining flexibility to adopt new models and technologies while ensuring that their platform can scale without crippling their budget. We're focused on empowering these teams with reliable, scalable open-source solutions that give them the freedom to innovate, knowing that their hybrid multicloud and/or on-premises data infrastructure is ready for whatever comes next."

To ground AI with your unique business context, vector databases have become essential. They allow you to store and to query high-dimensional vector embeddings that are generated from your data, enabling sophisticated semantic search and retrieval-augmented generation (RAG) applications.

The good news is that many open-source databases, including Cassandra, PostgreSQL and OpenSearch support vector datatypes and similarity search capabilities. That means you can avoid having to deploy a dedicated vector database to for AI workloads and instead you can use the open-source technology you already have.

We are proud to announce the expansion of NetApp enterprise-ready vector store capabilities. You can now build and manage powerful vector stores by using familiar, trusted open-source technologies like PostgreSQL using pg_vector. To maintain performance and reliability at scale, the Instaclustr Managed Platform integrates with leading storage providers, including:

• PostgreSQL on Azure NetApp Files
• PostgreSQL with Amazon FSx for NetApp ONTAP
• Coming soon: PostgreSQL with Google Cloud NetApp Volumes

By combining the flexibility of PostgreSQL with robust, high-performance NetApp storage, you can build a secure and scalable foundation for your AI applications. Your team can confidently ground AI models with specific domain knowledge so that your applications deliver accurate and relevant results while keeping your data secure.

Delivering a fast, intelligent, and responsive search experience is critical for the success of any AI application. Slow or irrelevant results can quickly frustrate users and undermine the value of your solution. To overcome this challenge, we are focused on helping you deploy highly efficient AI pipelines.

AI Search for OpenSearch introduces new capabilities that allow users to create a fully managed AI search pipeline in minutes. It does this by automatically deploying an ingest pipeline and machine learning model to generate embeddings stored in a vector index. It also introduces AI-based search that understands query intent beyond keywords. Instaclustr offers fully managed services and support for these AI capabilities backed by our world-class SLAs.

While AI unlocks new capabilities, it is critical that your data infrastructure remains performant and cost effective. This is why we are announcing upcoming support for GPU- nodes for OpenSearch. This new feature allows you to add a dedicated GPU instance to your OpenSearch cluster, supercharging processing of ML tasks without affecting overall cluster performance.

What does this capability mean for your team?

• Faster searches. Use the parallel processing power of GPUs to drastically reduce query latency for complex neural searches.
• High efficiency processing. Optimize tasks with GPU nodes to reduce cluster strain, minimize regular data node compute needs, and lower resource consumption.
• Simplified architecture. Reduce the complexity of your AI pipeline by running models closer to your data, minimizing data movement and potential points of failure.

This advancement is designed to give you a competitive edge, empowering you to build AI search experiences that are not only powerful but also incredibly efficient.

Modern businesses face relentless pressure to deliver real-time insights from rapidly expanding, distributed data sources. The integration of Instaclustr managed services for ClickHouse with Amazon FSx for NetApp ONTAP brings a major leap forward for organizations adopting a hybrid lakehouse architecture.

By uniting ClickHouse's high-performance analytics engine with FSx for ONTAP’s scalable, fully managed storage, teams can:

• Simplify data management. Query data directly from FSx for ONTAP, eliminating complex data pipelines and streamlining analytics.
• Unlock cost savings. Optimize spending with FSx for ONTAP tiered, S3-compatible storage, reducing costs while ensuring high availability.
• Drive real-time analytics. Store all data types (structured, semi-structured, and unstructured) in a single, secure repository for fast, direct queries, driving advanced analytics, BI, and AI/ML.

This integration empowers you to build a truly unified data lakehouse, accelerating insights and unlocking new possibilities, all while freeing your team to focus on data-driven innovation instead of infrastructure maintenance.

Why Top AI Innovators Trust ClickHouse

ClickHouse is the high-performance analytical database trusted by leading AI and big data companies for its unmatched speed, reliability, and scalability. Leaders like OpenAI and Anthropic leverage ClickHouse for petabyte-scale observability, ensuring their AI models are performant and reliable. This integration brings you the same advanced analytics engine trusted by the very forefront of AI—and the freedom to innovate without limits.

An effective AI application rarely relies on a single data source. It needs to pull information from streaming platforms, transactional databases, operational systems, and analytical data lakes. Integrating these disparate sources into a cohesive pipeline for AI agents and applications is often a complex and time-consuming task.

To solve this major pain point, NetApp is developing a groundbreaking solution: the Model Context Protocol (MCP) gateway. This unified gateway, coming soon to the Instaclustr platform, is designed to streamline the entire data integration process for AI.

The MCP gateway will act as a central, intelligent layer that simplifies how your AI applications and agents access data. It will provide a unified protocol for integration of information from streaming, transactional, operational, and analytical sources. This unified approach will significantly reduce the engineering effort that’s required to build and to maintain data pipelines, freeing up your team to focus on developing innovative AI features rather than wrestling with data plumbing.

The announcements from NetApp INSIGHT 2025 represent our deep commitment to providing you with a secure, scalable, and cost-effective, fully managed data platform on premises or across any cloud for your AI journey. By embracing open-source technologies, NetApp empowers you to build the flexible, powerful, and future-proof infrastructure that your business needs to succeed.

From enterprise-ready vector stores to GPU-accelerated AI pipelines and simplified data integration, our goal is to help you reduce operational costs, avoid vendor lock-in, and accelerate innovation. We are here to be your trusted partner, providing the expert support and managed solutions that you need to confidently navigate the evolving AI landscape.

To get started, learn more about NetApp Instaclustr managed open-source services.

Then take the next step toward optimizing your open-source infrastructure for AI and schedule your free consultation.