NetApp® announced it is collaborating with NVIDIA to advance retrieval-augmented generation (RAG) for generative AI applications. The new collaboration directly connects the just-announced NVIDIA NeMo Retriever microservices?coming to the NVIDIA AI Enterprise software platform for development and deployment of production-grade AI applications, including generative AI?to exabytes of data on NetApp?s intelligent data infrastructure. Every NetApp ONTAP® customer will now be able to seamlessly ?talk to their data?

to access proprietary business insights without having to compromise the security or privacy of their data. Enterprises want to leverage publicly available large language models to talk directly and securely to their corporate data with assurance that their private data will never be accessible outside the enterprise. Until now, enterprises that wanted to build internal chatbots, co-pilots, and applications that leverage corporate knowledge were faced with complexity and uncertainty about how to get started without compromising data security or privacy.

NetApp and NVIDIA have developed a simple solution with NVIDIA NeMo Retriever microservices technology for RAG that can leverage any data stored on NetApp ONTAP?on-premises and in the leading public clouds. NetApp customers can now query their data, whether spreadsheets, documents, presentations, technical drawings, images, meeting recordings, or even data from their ERP or CRM systems through simple prompts ? all while maintaining the access control they?ve already established when storing the data.

By combining NVIDIA?s NeMo Retriever microservices with NetApp ONTAP?s footprint, tens of thousands of enterprises both on-premises and in the world?s public cloudscan access their data wherever it resides. This reduces the friction, cost, and time to value for RAG. The new capability allowing customers to talk to their corporate data specifically for enterprise RAG complements NetApp?s robust portfolio of mature AI offerings, which have been leveraged by more than 500 of its joint customers for AI model training and inference, including solutions built on NVIDIA DGX BasePOD and that have certification for NVIDIA DGX SuperPOD, as well as the new NVIDIA OVX systems storage validation program, designed specifically for enterprise RAG.