This work introduces a prototype of a user-centered mobile tourism recommender system that integrates a simplified Multi-Criteria Decision Making (MCDM) method with an LLM-based natural language explanation interface to enhance the usability and transparency of sustainable tourism planning. The system allows users to weight four sustainability-related criteria - crowd levels, weather, air quality, and distance. The Borda count method is used to compute final rankings and recommendations based on real-time and simulated data. A built-in assistant, powered by OpenAI's GPT-3.5 model, explains recommendations through natural language interaction. A user study (N=13) using the System Usability Scale (SUS) and open-ended questions to reveal satisfactory usability, appreciation for the interactive interface, and perceived transparency of the system. Limitations are also discussed, including the lack of contextual guidance during initial criteria weighting and the assistant's vulnerability to off-topic queries, underscoring the need for improved prompt design and more structured user interaction.