Homelab · Mac Mini · Cloudflare Tunnel

LLM Full Stack Portfolio

Three live demos that answer one interview question each — RAG, agents, and evaluation. Built with Nuxt, FastAPI, Postgres + pgvector, and Ollama. $0 hosting.

Nuxt 3 FastAPI pgvector Ollama launchd

Track 1

PDF Chat

"Walk me through your RAG pipeline."

  • Upload & chunk PDFs
  • Semantic search
  • Source citations
Open RAG demo →

Track 2

Tool Agent

"When should an LLM use a tool vs answer directly?"

  • Calculator
  • Weather API
  • Job search
Open agent demo →

Track 3

AI Eval

"How do you know your AI got better?"

  • Test suites
  • Accuracy & latency
  • Cost per query
Open eval dashboard →