Not just a VPN.
A mesh.
VPMN is the private mesh networking layer of the Umbra-Noesis stack — peer-to-peer, zero-trust, and local-first. Your nodes connect directly. No centralized server to route through. No external infrastructure to depend on. Your network. Your rules.
What it is
Traditional VPNs route your traffic through a central server — which means there's a central server someone controls, logs, or can cut off. A mesh network has no center. Each node connects directly to the others. Remove any single node and the rest of the mesh continues routing around it.
VPMN applies that architecture to the Umbra-Noesis stack — linking PCe nodes, field units, home labs, and remote systems into a private fabric that you own and control. When it ships, it will be the layer that lets your local AI infrastructure communicate across locations without ever touching third-party routing.
Peer-to-peer — nodes talk directly, no relay required
Zero-trust architecture — every connection authenticated, nothing assumed
Resilient — no central point of failure, no single point of control
Umbra Link integrated — identity-aware routing across all nodes
Architecture Pillars
Mesh Topology
Every node is a peer. Traffic routes through the most efficient path available — not through a bottleneck. The network self-heals when nodes join or leave.
Zero-Trust Model
No node is trusted by default — not even your own. Every connection requires authentication. Access is scoped, logged, and revocable at the node level.
Identity-Aware Routing
Routing decisions carry Umbra Link identity context — the mesh knows not just which nodes exist, but who is operating them and what they're authorized to reach.
Network Independence
VPMN operates over any available transport — LAN, WAN, cellular, or satellite. The mesh adapts to whatever connectivity is present without requiring a dedicated circuit.
Use Cases
Home Lab
Link desktop, NAS, GPU rig, and PCe nodes into a single private fabric — accessible from anywhere on the mesh.
Field · Remote
Connect boats, trucks, and remote deployments back to the home lab over any available connection — satellite, cellular, or port WiFi.
AI Node Clusters
Distribute inference workloads across multiple PCe nodes on the mesh — coordinated without a cloud backend or third-party orchestration layer.
Licensed Deployments
Operators building on the Umbra-Noesis stack get VPMN as the networking backbone — private mesh connectivity between all deployed nodes.
Roadmap
The architecture is complete. The implementation is in active development inside the Umbra-Noesis stack. VPMN will ship when it meets the same standard as every other layer — deterministic, governed, and production-ready.
Architecture & Design
Mesh topology, zero-trust model, identity binding with Umbra Link, and transport independence — all defined.
CompleteCore Implementation
Node discovery, encrypted peer connections, and routing logic — active development inside the EVE OS environment.
In DevelopmentEVE OS Integration & Release
VPMN ships as part of the EVE OS stack — validated against PCe hardware profiles and AACP protocol governance before public release.
RoadmapStack Position
Umbra Link
Provides the identity layer VPMN routes through — Node IDs and User IDs travel with every mesh connection.
EVE OS
VPMN runs as a system-level service on EVE OS — same as EVAA Cognition. The mesh is infrastructure, not an app.
PCe Systems
Every PCe node is a VPMN participant. The hardware tier determines the node's networking capacity and bandwidth contribution to the mesh.
The principle
Your network.
Your rules.
No one else's infrastructure.
VPMN is in active development. If you're building on the Umbra-Noesis stack and want to be informed when it ships — or want to discuss early access — reach out directly.