The Future Belongs To You

About the Role

We are building our own AI brain at SKXYWTF LLC (World Trade Factory), an AI-native financial intelligence company running SKORE—an institutional-grade AI equity research pipeline. We are moving inference off third-party APIs to our own GPU infrastructure in Phoenix, AZ.

Responsibilities

• Sets up and owns the WTFXAI inference server hardware powering the entire AI stack.
• Configures a GPU inference server running vLLM and Ollama.
• Quantizes open-weight models (Llama 3.2, Mistral, Phi-4, FinBERT) for optimal VRAM efficiency.
• Exposes a unified OpenAI-compatible inference API via Cloudflare Tunnel.
• Wires the server into our AI router to route all open-weight calls locally.
• Buils monitoring dashboards tracking GPU utilization, VRAM, temperature, and cost savings.
• Produces reports comparing API costs vs local inference at scale.

Requirements

• Linux systems administration (Ubuntu Server)
• CUDA and GPU optimization basics
• Python proficiency
• Networking fundamentals (ports, tunnels, reverse proxy)
• C++ or parallel computing experience is a strong plus.

This role offers remote-first work with relocation to Phoenix, AZ for server setup covered by the company. It starts as an internship with a clear path to full-time employment.