Why a sovereign cloud

When the institutions you build are intended to outlast you, you cannot rent your foundations. Osage Cloud runs on infrastructure the ecosystem owns and operates — Kubernetes clusters under our own management, owned container registries, owned ingress — on top of Hanzo PaaS. We pay for compute. We do not pay for permission.

What we run

• Compute. Kubernetes via Hanzo PaaS — DOKS in U.S. East, GKE in U.S. Central and EU-West-1, with reserved capacity in Singapore for Asian-hours latency targets.
• The 36-site fleet. Every osage.* property runs as an independent Cloudflare Pages project fed by its own per-site repo. Source-of-truth is decentralized; deployment is uniform.
• The news ticker worker. news.osage.network proxies six public RSS feeds (BBC, Reuters, Al Jazeera, AP, NPR, DW) and returns CORS-enabled JSON for the ticker on every site.
• Edge and ingress. hanzoai/ingress with the static plugin and gateway. White-labeled by hostname — the Hanzo, Osage, Lux, and Zoo brands serve from the same chassis without bleeding.
• Container registries. GHCR with per-org namespaces (ghcr.io/osagebrothers/*, ghcr.io/osagedao/*). No cross-org pulls.
• Builds. Native amd64 and arm64 in parallel on owned runners. No QEMU emulation. No GitHub-hosted runners for anything where build time matters.
• Secrets. Hanzo KMS only. Plaintext credentials do not exist in this ecosystem.

Regions

• U.S. East (primary). Reston / Ashburn region.
• U.S. Central (failover). Dallas region.
• EU West. Frankfurt for European partners and EU data residency.
• Asia Pacific. Singapore for Asian-hours latency and Asia-Pacific data residency.

Availability and SLO

Standing target: 99.95% monthly availability for all public-facing properties. Internal services run on the same chassis with the same SLO. Incidents are posted publicly with root-cause analysis within five business days. We do not run a status-page theatre — if a service is down, the dashboard shows it down.

What we do not run

• nginx. Caddy.
• Vendor-locked managed services for anything we can reasonably operate ourselves.
• Multi-tenant SaaS where the contract permits the vendor to evict us.

Pricing posture

Osage Cloud is not in the business of being the cheapest cloud. Internal workloads run at cost. Partner workloads (tribal nations, aligned institutions) co-locate at the internal rate plus a small operating margin. We do not run loss-leader pricing to acquire third-party customers; that path leads to either a margin we cannot defend or a service we cannot operate.

Data residency

Customer and partner data resides in the region designated at onboarding. Cross-region replication for disaster recovery stays within the same jurisdictional regime. No customer data is mirrored outside the regimes the customer selected.

For partners

Tribal nations and aligned institutions who require sovereign hosting can co-locate workloads on the Osage Cloud at cost, with their own brand on the same chassis. Inquiries go through osage.tech.

Engagement

Hosting inquiries: hosting@osage.cloud. Incidents and on-call: oncall@osage.cloud. Security: security@osage.cloud.