I will set up a ai or mcp server on your own hardware


About this gig
Your own AI server, on your own hardware no per-token bill, no data leaving your network.
WHAT YOU GET
- A model server installed and running on your machine: Ollama for simplicity, vLLM for throughput
- Web chat UI plus an OpenAI-compatible API endpoint, so existing tools can point at it
- An MCP server configured and connected to your real tools (files, git, databases, browser)
- Auth and network lockdown: reverse proxy, API keys, firewall rules, optional VPN not an open port
- Written setup notes so you or your IT team can restart and maintain it
WHY ME
I'm a network engineer with 10+ years running carrier-grade servers and networks, and I run this exact stack myself: Docker on a NAS, a multi-provider model routing gateway, n8n automation, and MCP tools. I build it on your box the same way.
The part most people get wrong isn't installing Ollama in one line it's access control and network isolation. That's my day job.
WHAT I DON'T DO
- I don't sell hardware or rent you compute
- I don't promise model intelligence that's the model's job, not the setup's
- Model training and fine-tuning are quoted separately
Not sure what your hardware can run? Message me before ordering
Get to know Yuri
I will fix your network issues and configure your router, switch and VLAN
- FromChina
- Member sinceDec 2022
- Avg. response time1 hour
Languages
English, Chinese
