VM provisioning
Shared CPU and Dedicated CPU instances across Sofia and Amsterdam, managed from our platform panel or API
RKE2 Kubernetes clusters
Automated provisioning with kubectl access through HAProxy, workload load balancing via external Caddy — cluster nodes remain on private networks with no public IPs exposed
Managed Kubernetes for teams not familiar with Kubernetes
We handle ArgoCD, Helm deployments, image updates, Vault secrets, and Ceph persistent storage. Your team only needs a Dockerfile and a GitLab CI pipeline to build your container image — we handle the rest
Geo-distributed application topology
Independent compute clusters per region (Kubernetes or VM-based) coordinated at DNS and database layer, with Caddy load balancers routing across both transparently
Cross-region master-master DB replication
MySQL/MariaDB with advisory on schema design to keep replication healthy under load
Regional failover automation
One-click tenant switching via Cloudflare API, including scheduler migration to the active region
Zero-downtime migrations using parallel subregion deployments
Migrate tenants one by one, roll back individually if needed
Cross-region cache replication
Redis/Dynomite clusters without Redis Enterprise licensing costs
Observability stack
Prometheus and Grafana with dynamic target discovery via prom-sd-api, supporting a growing set of exporters: node, MySQL, PHP-FPM, MikroTik, iLO, Proxmox — new monitored targets added without cluster changes
Secrets management
Vault with auto-unseal, ExternalSecrets operator, per-service secret deployment managed through the platform panel
Private L2/L3 networking
Per-client isolated networks across Sofia and Amsterdam, with full cross-region VPN available for clients requiring it
Colocation in Sofia
Rack space in Telepoint/Digital Realty and SDC2/Neterra, remote hands, BGP connectivity, and optional L2 integration between colocated physical servers and Radibase cloud VMs
We work with SaaS companies and engineering teams that need more than a cloud provider offers — dedicated infrastructure, hands-on ownership, and engineers who actually know their stack.