"Netrunner team on AMD AI League Hackathon"

Team Idea

Building a full-stack AI web application using FastAPI, React, and AMD ROCm for optimized model inference. I am participating solo for this challenge, so I only need feedback and technical insights from the community and mentors.

Team Members

Unknown Hacker

adityasuryavanshidev91062

Submission

PHANTOM

PHANTOM is a hardware-transcendent local LLM runtime. It orchestrates VRAM, system RAM, and NVMe as a single memory tier — so a 6 GB laptop GPU can run 70B models at conversational speed.