Pillar 1 — Cloud
A sovereign cloud, on your terms
The building blocks of the big clouds — compute, networking, S3-compatible storage, secrets, registry, monitoring — on infrastructure you control. Production-grade by default: a redundant 3-node, multi-site, replicated stack. Buy it fully managed, or license it and run your own cloud on your own hardware — from a single app to an enterprise fleet.
- Compute & networking
- Proxmox virtual machines on software-defined MikroTik networks — isolated per project.
- Object storage
- S3-compatible MinIO, with optional replication and off-site backup.
- Secrets & identity
- HashiCorp Vault for every credential — nothing hard-coded, nothing shared.
- Private registry
- Harbor namespaces per project; your container images never leave the platform.
- Ingress & TLS
- Traefik routing with automatic certificates — public or internal-only.
- Monitoring & backups
- Health, metrics and scheduled backups wired in from day one.
This is not a cheap VPS. It is a sovereign, production-grade cloud — the same stack ATAILA runs in production, including the AI models serving your apps — and it's typically cheaper than a hyperscaler, not pricier.
The SLA
A number in writing: 99.9%.
Managed production runs under a written SLA — 99.9% availability, with recovery time (RTO) and recovery point (RPO) defined per plan. It's in the contract, not the footer — and it's backed by the redundant, multi-site, monitored stack above, not by a promise. When you compare offers, ask every vendor the same question: what infrastructure stands behind your uptime figure?
How pricing & the SLA work →Two ways to run it
Fully managed on ATAILA, or licensed on your own hardware — single-tenant, your jurisdiction, still operated by us. Any size.