Travel Planner Chatbot | Python, Transformers, HuggingFace, FAISS, Streamlit, SpeechRecognition
• Developed an AI-powered chatbot for travel planning using Retrieval-Augmented Generation (RAG), combining
speech-to-text, semantic search, and large language models.
• Integrated Qwen2.5 LLM fine-tuned on the OSU TravelPlanner dataset with a FAISS vector database for efficient
information retrieval.
• Implemented voice input with SpeechRecognition API and built an interactive Streamlit UI for real-time queries.
• Enabled the chatbot to generate multi-day itineraries with explanations (justifications) for each recommendation.