Building My Own Cloud HPC Setup for AI Research

I'm slowly putting together a personal cloud-based HPC setup to support my AI research journey. Right now it's simple: a GPU machine from Lambda, a CPU box from DigitalOcean, automatic environment bootstrapping, and persistent storage via a Git repo so I can spin up fresh instances anytime and keep working.

It's not polished yet—no unified interface, no shared filesystem, and only a test environment—but it's my first real step toward owning my research stack and eventually building my own lightweight scheduling system.