Created a dynamic itinerary generator using multi-head genetic algorithms and RAG from YouTube and Instagram data.
This innovative travel planning system creates personalized itineraries by analyzing social media content from platforms like YouTube and Instagram to identify trending destinations and authentic experiences.
Using a multi-head genetic algorithm, the system optimizes travel routes based on user preferences, budget constraints, seasonal factors, and popularity metrics.
The Retrieval Augmented Generation (RAG) component ensures recommendations include up-to-date information about attractions, local events, and hidden gems not typically found in traditional travel guides.
Implemented a hybrid data approach combining vetted databases with real-time social scraping and verification through a multi-source consensus algorithm.
Developed a progressive preference learning system that refines recommendations based on user interactions and explicit feedback.
Utilized an advanced constraint solver alongside the genetic algorithm to handle complex interdependencies between destinations, timing, and budget.
"The AI recommendations uncovered amazing places I would have never found through traditional research. My Tokyo trip had the perfect balance of popular attractions and local secrets."
Emma Rodriguez
Travel Enthusiast & Beta Tester
