Aranya raises $11M to turn bare-metal servers into AI clusters in 48 hours

Aranya's pitch is that AI inference infrastructure is messy. Real GPU hardware is heterogeneous, and existing orchestration tools like Kubernetes are not designed to handle hardware-level failures. The company claims clusterdOS can absorb that complexity, turning racked servers into a self-healing cluster within 48 hours, a promise it says no other operator consistently matches for custom architecture.

The startup was founded last year and already manages over $500 million in GPU hardware, which suggests early traction with AI inference providers and data centers. The funding, split between a $9 million seed and a $2 million pre-seed, will go toward expanding engineering and sales teams and launching a full multicluster interface. The focus on federated control and utilization monitoring aims to cut downtime and operating costs, which could appeal to organizations running AI workloads at scale.