LLM Full Stack Portfolio — three live demos on a Mac Mini — Blog

The north star for this build was simple: become credible for an LLM Full Stack Engineer role with portfolio proof — not by studying everything, but by shipping three projects that each answer one interview question.

They're all live now at lab.uft1.com.

The three tracks

Track 1 — RAG (rag.uft1.com)

Interview question: "Walk me through your RAG pipeline."

Upload a PDF → chunk → embed with OpenAI → store in Postgres + pgvector → ask questions → answer with source citations.

Stack: Nuxt 3 UI, FastAPI API, pgvector, Ollama for local chat.

Track 2 — Tool Agent (agent.uft1.com)

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

An agent loop that picks tools when the model needs live data, precision math, or external APIs. Three tools shipped: calculator (AST-safe eval), weather (wttr.in), and job search (local JSON).

The UI shows the full tool trace — every call, args, and result — so you can see the orchestration, not just the final answer.

Track 3 — AI Eval (eval.uft1.com)

Interview question: "How do you know your AI got better?"

Fixed test suites that fire questions against the live RAG and agent APIs, then score accuracy (keyword + tool-call matching), latency, and cost. Run a suite, get a before/after baseline — that's how you prove a retrieval tweak or model swap actually helped.

Infrastructure ($0 hosting)

Internet
  → Cloudflare Tunnel (mac-mini-local)
  → localhost:PORT (launchd)
  → Postgres / Ollama / FastAPI / Nuxt

Piece	Choice
Inference	Ollama on M1 — free
Embeddings	OpenAI `text-embedding-3-small` — pennies
Database	Postgres + pgvector via Homebrew
Deploy	rsync to boot volume + launchd plists
Public HTTPS	`*.uft1.com` on Cloudflare Tunnel

No Docker, no AWS, no monthly bill. I already had homelab experience — this was about proving I can wire AI apps end-to-end on real infra.

What I'd say in an interview

I built a RAG app with citations over pgvector, a tool-calling agent with visible orchestration, and an eval harness that benchmarks both with fixed test suites. All three are live demos on my homelab, exposed through Cloudflare Tunnel.

Three sentences. Three URLs. One story.

Jovylle Bermudez