Autonomous AI agent system that automates over 90% of the travel planning process using CrewAI, Gemini API, and Streamlit.
The Intelligent Travel Planning System is an end-to-end autonomous agent application that automates over 90% of the travel planning process — from destination research and itinerary generation to accommodation recommendations and budget estimation.
Three specialized CrewAI agents collaborate: a Research Agent gathers real-time destination information, a Planner Agent structures the itinerary based on user preferences, and a Coordinator Agent assembles the final plan and handles follow-up questions. The system is powered by the Gemini API for its long-context reasoning capabilities.
A Streamlit conversational interface allows users to express their travel preferences naturally. The application is integrated with a REST API and deployed via GitHub Actions CI/CD, making it continuously updated and production-grade.
Understanding agents, tasks, tools, and crew orchestration patterns.
Defining the Research, Planner, and Coordinator agents with precise roles, goals, and tool access.
Building custom tools: web search wrapper, destination data fetcher, budget calculator.
Configuring CrewAI to use Gemini API with proper context window management for long plans.
Building a multi-turn conversational interface that displays agent reasoning steps in real time.
Writing GitHub Actions workflows for automated testing and Streamlit Cloud / cloud deployment.