The collaboration combines NVIDIA AI retrieval technologies with Commvault Data Activate to make enterprise data accessible to AI agents while preserving permissions, governance, data lineage and security.
The next phase of enterprise AI will depend less on the sophistication of AI models and more on the quality, accessibility and governance of the data feeding them. Recognising this critical requirement, Commvault and NVIDIA are collaborating to address one of the biggest barriers to enterprise AI adoption: making organisational data usable by AI applications and autonomous agents without compromising security and control.
The integration brings together NVIDIA NIM microservices, NVIDIA NemotronRetriever and Commvault Data Activate, enabling enterprises to extract, organise and make Commvault-protected information accessible to AI-driven workflows.
The announcement comes at a time when organisations across the Middle East are moving from AI experimentation towards broader enterprise deployment. According to Deloitte’s 2026 State of AI in the Enterprise report, 66% of Middle East organisations report improvements in efficiency and productivity through AI, but only 34% are using it to fundamentally transform their products, processes or business models. This gap highlights the challenge of translating AI investments into deeper business transformation.
At the heart of the collaboration is the recognition that enterprise AI cannot operate effectively on fragmented, outdated or poorly governed information. As organisations deploy increasingly autonomous agents, ensuring that these systems access accurate, authorised and current data becomes essential.
Using NVIDIA NIM Nemotron Retriever and NVIDIA cuVS with Apache Solr, the reference design will enable organisations to make complex enterprise content easier for AI applications and agents to retrieve and process.
“Enterprise agents need current, trusted data to turn business knowledge into action.” — Pat Lee, Vice President, Strategic Partnerships, NVIDIA
An important capability is the continuous synchronisation of information available to AI systems. Through Data Activate, changes in source data, deletions, expired retention periods and modifications to access permissions can be reflected in the information made available to AI. The integration also preserves data ownership, classification and lineage, helping enterprises understand where information originates, how it is used and who can access it.
For CISOs, this addresses a critical emerging risk: AI agents potentially accessing information beyond their authorised scope or continuing to use data whose permissions have changed. Maintaining governance throughout the AI data pipeline could become a fundamental requirement as autonomous systems gain access to sensitive enterprise information.
“The Middle East is making significant strides in advancing its AI ambitions, creating new opportunities for businesses to innovate and scale,” says Fady Richmany, Corporate Vice President and General Manager, Emerging Markets, Commvault.
NVIDIA’s Pat Lee, Vice President, Strategic Partnerships, emphasises the importance of trusted information in enabling enterprise agents to deliver meaningful outcomes.
Strategically, the collaboration reflects an important evolution in enterprise data management. Data protection is no longer limited to backup, recovery and business continuity; it is increasingly becoming an enabler of secure AI adoption. Organisations must ensure that protected data can also be activated, accessed and governed across emerging AI ecosystems.
For CIOs, the opportunity lies in extracting greater business value from existing enterprise information rather than relying exclusively on new data infrastructure investments. For CISOs, the priority is ensuring that increased data accessibility does not translate into uncontrolled access.
The SaaS and software reference design is targeted for Early Release in December 2026, with general availability planned for 2027. The companies will provide further insights into Data Activate at SHIFT Virtual 2026 on November 18.
