DevOps Engineer to Set Up Private AI Agent Server (One-Time Setup & Training Only)

I am looking for an experienced DevOps / automation engineer to set up a secure, private AI-powered automation system on my own VPS. This is a setup and training project only; I am not looking for long-term management. The goal is to build a clean, scalable foundation and train me to operate and extend it myself.

My current server:

Ubuntu 24.04 VPS

2 CPU cores, 8 GB RAM

Full root access

Phase 1 goal:

Secure Linux server setup with Docker (no Kubernetes)

Modular Python-based agent framework

One main agent that can delegate structured tasks to at least one sub-agent

LLM integration (Claude or similar, via API)

Lightweight web interface to submit natural language instructions

Structured JSON output layer

Logging and basic scheduled task capability

Clear documentation and knowledge transfer (so I can operate it myself)

Important constraints:

No Kubernetes

No browser automation

No automatic live pushing of changes to production

Clean, modular, maintainable architecture

Focus only on a solid V1 foundation

When you apply, please include:

Your proposed high-level architecture

Estimated total hours for Phase 1

Your proposed fixed price for Phase 1 (even though this is hourly)

Estimated timeline

1–3 examples of similar systems you have built (AI agents, automation platforms, private LLM setups, etc.)

Please only apply if you have real experience with DevOps, Docker, Python-based backends, and integrating LLM APIs. I am ready to move quickly once we are aligned on the scope.