Spinifex

Spinifex is an open-source platform that runs AWS-compatible APIs on infrastructure you own. If your software already talks to EC2, S3, EBS, VPC, or IAM, it keeps working on Spinifex without any code changes. Same AWS SDKs, same Terraform configs, same CLI. You change where the workload runs, not what it is.

The core use case is straightforward: move workloads off hyperscalers, run them on your own racks, at the edge, or on a Neocloud partner, without re-platforming. It's particularly relevant for sectors where data sovereignty, air-gapped deployments, or cost control matter most: defence, healthcare, financial services, energy, and AI/ML teams paying GPU rates to a hyperscaler.

Spinifex didn't just wrap an existing API layer. The engineering underneath was rebuilt from scratch:

• Object storage (S3-compatible): Reed-Solomon erasure coding across nodes for durability without triple-replication costs.
• Block storage (EBS-compatible): Volumes that keep serving when a node drops. Snapshots, attach/detach, and growable volumes included.
• Compute (EC2-compatible): QEMU/KVM on bare metal you own. No hypervisor tax eating into performance.
• Networking (VPC): Built on OVN/OVS.
• Identity (IAM): SigV4 authentication and policy enforcement.
• Kubernetes (EKS): Conformant K8s included in the same stack.
• Container registry and service (ECR/ECS): Available alongside load balancing (ALB/NLB) and managed databases (RDS).

Deployment shapes are flexible. Spinifex runs on a single node, a cluster, an edge box, or an air-gapped rack. The API surface stays the same across all three.

Licensed under AGPL-3.0, the core is auditable and self-hostable. You're not dependent on a vendor's continued goodwill to keep running your own infrastructure.