RAVEL x RT-One: sovereign AI moves to multi-site orchestration
RAVEL's 23 June 2026 release says RT-One will use RAVEL Orchestration to federate sovereign data centers across Brazil and the United States, with an initial scope announced at 1 GW of capacity.
1. What is officially announced
RAVEL says RT-One selected its orchestration platform to coordinate distributed sovereign AI infrastructure across the Americas. The release cites an initial 1 GW footprint across Brazil and the United States, aiming to unify compute, power, and data-center resources into an operating model for large AI workloads.
2. Why this is a sovereign AI signal
The important point is not only the amount of capacity. The signal is the operating model: sovereignty is no longer limited to hosting a model or dataset in one country. It also depends on orchestrating workloads across several sites while preserving jurisdictional boundaries, governance rules, and controlled power constraints.
3. Operational takeaway for organizations
For AI Belgium, AI France, Odoo Belgium, Odoo France, and Odoo Enterprise programs, this news underlines the need to design federation from the start: data location, workload segmentation, exit policy, multi-site supervision, and the ability to move or retain processing according to regulatory constraints.
Map AI workloads by criticality, jurisdiction, and GPU demand before choosing a sovereign multi-site orchestration strategy.
Start scoping