Cloudian has announced the general availability of the Cloudian HyperScale AI Data Platform, designed to enable organisations to convert decades of institutional knowledge into immediate, AI-driven insights. The launch is based on the NVIDIA AI Data Platform reference design and provides an on-premises AI solution aimed at businesses seeking complete data sovereignty.
The platform is intended to solve a longstanding challenge for enterprises, many of which have struggled to adopt AI without compromising compliance or embarking on lengthy, complex development projects. A significant proportion of enterprise knowledge remains locked in unstructured files, including documents, imagery and operational records that AI systems are often unable to access.
Cloudian’s HyperScale platform combines compute, storage and software within a single integrated system, using NVIDIA AI Blueprints to support enterprise document AI. Unlike bespoke deployments, the solution is designed to be operational within days and requires minimal data migration.
“The Cloudian HyperScale AI Data Platform is built on NVIDIA AI infrastructure, networking and software to help organizations access critical knowledge securely.”
Justin Boitano, Vice President, Enterprise AI Products, NVIDIA
According to Cloudian, the platform eliminates traditional performance bottlenecks through S3-native object storage and RDMA for S3 performance, providing up to eight times faster vector database operations compared with CPU-based systems using file storage. The integrated appliance model also aims to remove unpredictable cloud expenses by removing per-token AI charges and high-cost storage fees, while ensuring all data remains within the organisation’s control.
Cloudian says the platform enables institutions such as hospitals and financial services providers to use AI to search and analyse confidential material without exposing information to external cloud providers. Healthcare organisations can use natural language queries to access records while maintaining HIPAA compliance, while regulated industries can interrogate documents and contracts securely.
“Organisations need AI that works with their existing security and compliance requirements, not against them. Our HyperScale AI Data Platform delivers exactly that: secure access to institutional knowledge through a familiar chatbot interface, with complete data sovereignty and deployment in days rather than months.”
Michael Tso, CEO, Cloudian
The company plans to expand capabilities beyond document AI with new blueprints for video analysis, covering industrial environments such as factories, warehouses and construction sites. Additional analytics and enterprise AI features will be deployed using the same infrastructure.
The entry-level system is built on the NVIDIA RTX PRO Server, incorporating up to eight NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, dual Intel Xeon 6900 CPUs, NVIDIA BlueField-3 DPU, NVIDIA Spectrum-X Ethernet SuperNICs, NVIDIA AI Enterprise software and Cloudian HyperStore NVMe-based storage. The platform supports RDMA for S3 connectivity and can scale from terabytes to exabytes.
“Turning business data into actionable AI insights requires more than just storage – it demands integrated, high-performance infrastructure that evolves with enterprise needs.The Cloudian HyperScale AI Data Platform is built on NVIDIA AI infrastructure, networking and software to help organisations access critical knowledge securely, reduce deployment complexity and scale their AI capabilities as their data grows.”
Justin Boitano, Vice President, Enterprise AI Products, NVIDIA
The HyperScale platform is now available worldwide through Cloudian’s certified channel partners. Each deployment includes Cloudian HyperCare support with remote management and round-the-clock technical assistance. Air-gapped configurations can also be deployed for organisations requiring maximum isolation from the cloud.
