Edgeless Systems

Edgeless Systems builds Constellation, an open-source Kubernetes distribution that runs entire cluster workloads, including the control plane, inside hardware-based confidential computing enclaves (AMD SEV-SNP and Intel TDX). The effect is that node memory is encrypted at the hardware level even from the cloud provider’s own hypervisor, extending the isolation promise of confidential computing from individual processes to a full orchestrated cluster — a materially harder engineering problem than protecting a single VM.

This targets a specific and growing anxiety among regulated and security-sensitive cloud customers: that a cloud provider’s own privileged access, or a compromised hypervisor, represents a residual trust boundary that standard encryption-at-rest and in-transit does not address. Confidential computing closes that gap by keeping data encrypted in use, and Edgeless Systems’ contribution is making that guarantee operable at the scale of a real Kubernetes deployment rather than a single isolated enclave, alongside a companion product line for confidential AI workloads that extends the same protections to LLM inference and training.

Founded in 2020 and headquartered in Bochum, Germany, Edgeless Systems has raised a seed round of $5 million (2023) from investors including Berlin’s SquareOne, on top of earlier funding, bringing disclosed funding to roughly $6.7 million. Constellation itself is open source, which gives outside security researchers the ability to inspect the implementation directly rather than relying on vendor claims — a real efficacy advantage over closed-source confidential computing offerings, even though the company’s commercial scale remains small relative to major cloud providers’ own confidential computing services.