30/07/2026
⚙️ How to run Rocky AI on your own VPS
Want Rocky AI running on your own server instead of waiting on the public launch? Here's the general flow:
1. Pick your VPS
Minimum: 4 vCPU, 8GB RAM, Ubuntu 22.04+. More RAM if you're running multiple agents concurrently.
2. Base setup
sudo apt update && sudo apt upgrade -y
sudo apt install docker.io docker-compose -y
3. Environment
Set up your API keys, DB (Postgres + pgvector for memory), and Redis for task queuing — all containerized so nothing touches the host directly.
4. Deploy Rocky AI
Pull the service, configure your agent routing (customer, DevOps, security), and bring it up behind a reverse proxy (Nginx/Caddy) with SSL.
5. Point it at Rocky Agentic OS (optional)
If you're running the full stack, Rocky AI registers into the OS's integration layer — one API call, and it's part of the agent pool.
6. Monitor & scale
Watch resource usage per agent. Scale horizontally by adding VPS nodes behind a load balancer as usage grows.
Full 24/7 AI, running on infrastructure you control. No black box.
Details on setup access coming closer to the August 2026 launch.
