Cuon will launch the on-premises Private AI Appliance solution on 1 November 2026. It uses a two-tier architecture that enables secure operation of local LLMs while maintaining data sovereignty within the customer environment.
Cuon, in collaboration with Cuon of the TerraSky Group, will launch the on-premises Private AI Appliance solution, which enables the secure use of local LLMs through a two-tier architecture comprising Box installed in the customer environment and a management cloud. The service is scheduled to launch on 1 November 2026 (Sunday).
Designed with an emphasis on data sovereignty, the service sandboxes the highly confidential LLM runtime environment within the internal network, while authentication credentials used when connecting to external SaaS applications are securely managed and relayed through a gateway on the cloud side. The two-tier Box and Cuon Cloud configuration provides access to management functions such as operational monitoring and contract management.
The solution is provided in a two-tier configuration comprising Box and Cuon Cloud. Box is a server appliance installed within the customer network, while Cuon Cloud provides management functions. Communication between the two tiers is completed solely through outbound connections from Box, so opening inbound ports to the customer network or using a VPN is not required.
The support areas cited include Salesforce and DX consulting, CDP/DMP/security, and AWS, with an emphasis on the ability to support customers with complex, customer-specific environments.