Budget Constrained Travel Itinerary Planning Agent

Can a small local language model plan a multi-day trip that meets a budget, room rules, cuisine wishes and common sense all at once? This project builds a CAMEL-AI agent: a planner reads the reference data through search tools and drafts a day plan, a deterministic checker lists the violations, and a critic fills the plan slots they implicate, for at most two rounds. The model (qwen2.5:3b (8k copy)) runs in local Ollama on a laptop with RTX 2050 4 GB (Ryzen 7, 8 GB RAM).

NOTE: Research / teaching demo on the public TravelPlanner benchmark. Prices and options are benchmark values, not live offers. Nothing is ever booked or paid; the booking tool is a written-not-run stub with human approval.

Category: Department of Computer Science and Engineering / Agent Development

Agent Development Project AG15 from ProTechSC Academy, with graphical abstract, video demonstration and project guidance.