ALBERT NAHAS

Project One-Pager

Booker Agent

AI reservation workflows with an API

Open Sourceopen source

API-driven AI restaurant reservation agent.

Booker Agent

No screenshot attached for this project yet.

Why I Built It

The Meta Story

Booker Agent came from a simple question: can AI reliably handle small, repetitive real-world chores like restaurant booking without becoming brittle?

I picked a narrow task on purpose because constrained workflows are where useful agents should prove themselves first.

Building it taught me how to balance model intelligence with deterministic API logic and explicit fallbacks. The motivation was practical autonomy, not novelty.

What This Project Taught Me

  • Bounded scope makes agent reliability measurable.
  • Fallback logic is essential for real-world automation.
  • Useful autonomy starts with repetitive tasks people already dislike.