A vision for AI · cloud & local

AI, everywhere it runs.

One node-management system spanning the cloud, the local machine, and hardware at the edge — configured and connected as a single fleet.

Cloud
managed centrally
Local Machine
adapts on-device
Hardware
edge & accelerators
One system

A node-management system for all kinds of devices.

Every machine joins the same fleet and answers to the same control plane — no matter the form factor or where it lives.

Laptops
dev & desktop
Servers
on-prem & cloud
GPU rigs
accelerators
USB sticks
boot-to-AI
Configuration

Configured centrally. Adapted locally.

Set policy once for the whole fleet, then let any node override what it needs — for maximum adaptation to the hardware in front of it.

Central policy · plan.ai cloud
global
IDENTITY
openrouterCLOUD
hermesAGENT
ollama
PUSH
OVERRIDE
Local override · laptop
global
inherited
openrouter
offline
hermesAGENT
ollamaLOCAL
Agents

Works with every agent.

First-class support for the agents your teams already run — openclaw, hermes, opencode — plus anything you build yourself.

openclaw
autonomous coding
hermes
enabled
opencode
terminal-native
Bring your own
drop-in adapters
plan.ai relay

Reach any node, from anywhere.

Remote connectivity through plan.ai relay reaches devices behind NAT and firewalls — no inbound ports, no VPN, no static IPs.

Operator
anywhere
plan.ai relay
secure tunnel
FIREWALL / NAT
Node
behind firewall
relay shell — laptop
$ planai relay connect laptop
✓ tunnel established · 0 inbound ports
user@laptop:~/usb$
plan.aimgmt

Cloud-managed. Locally adaptive. Everywhere connected.