Inspiration
The inspiration for Hotel Win Win came from experiencing the frustrations of traditional hotel booking platforms - long response times for customer queries, delayed review approvals, and lack of intelligent assistance. We envisioned a modern solution that combines the convenience of online booking with AI-powered customer support and automated review management, creating a win-win situation for both hotel owners and guests.
What it does
Hotel Win Win is a comprehensive MERN stack hotel booking system that revolutionizes the hospitality industry through intelligent automation: For Users:
✅ Seamless room booking with real-time availability 🤖 Interactive AI chatbot for instant hotel-related queries 🔐 Google OAuth integration for quick sign-up/login ⭐ Review and rating system with instant feedback 🏨 Complete facility browsing and room details
For Admins:
📊 Comprehensive booking management dashboard 🏠 Dynamic room inventory control (add, update, delete rooms) 👥 User management with spam user blocking capabilities 🔍 Intelligent review moderation system 📈 Analytics and booking insights
AI-Powered Features:
🧠 Smart Chatbot: Powered by TinyLlama-1.1B-Chat model, trained on hotel-specific requirements using RAG (Retrieval-Augmented Generation) ⚡ Automated Review Approval: Uses sentiment analysis to automatically approve positive reviews and flag negative ones for admin review
How we built it
Architecture & Technology Stack Frontend: React.js deployed on Vercel Backend: Node.js/Express.js deployed on Render Database: MongoDB for data persistence Authentication: JWT tokens with Google OAuth integration File Storage: Supabase for image management
AI Implementation
Chatbot: TinyLlama-1.1B-Chat-v1.0 model (2GB VRAM, 4GB RAM) Training: Custom RAG implementation using hotel-specific PDF documents Deployment: Dockerized and hosted on Hugging Face Hub Review Analysis: Custom sentiment analysis model hosted on GitHub
Development Process
Planning: Designed user flow and admin workflow Backend Development: Built RESTful APIs with JWT authentication Frontend Development: Created responsive React components AI Integration: Trained and deployed the chatbot model Automation: Implemented sentiment-based review approval Testing & Deployment: Comprehensive testing and cloud deployment
Challenges we ran into
Technical Challenges Model Optimization: Balancing chatbot response quality with resource constraints (2GB VRAM limit) RAG Implementation: Training the model effectively with hotel-specific documents while maintaining response accuracy Real-time Performance: Managing chatbot response delays due to Hugging Face hosting limitations Cross-platform Authentication: Seamlessly integrating JWT with Google OAuth
Infrastructure Challenges Resource Management: Optimizing the TinyLlama model for limited cloud resources Docker Deployment: Containerizing the AI model for Hugging Face deployment API Rate Limits: Managing external service limitations across multiple platforms
Data Management Image Optimization: Implementing efficient image storage and retrieval with Supabase Review Processing: Creating accurate sentiment analysis for diverse review types
Accomplishments that we're proud of
Full-Stack AI Integration: Successfully implemented RAG-based chatbot in a production MERN application Intelligent Automation: Created a sentiment analysis system that reduces admin workload by 70% Scalable Architecture: Built a system that handles multiple users with real-time booking capabilities User Experience: Achieved seamless integration between traditional booking features and AI assistance Cost-Effective AI: Deployed a functional AI system using lightweight models and free-tier services Modern Authentication: Implemented secure JWT + Google OAuth system
What we learned
Technical Skills
AI Model Deployment: Learned to deploy and optimize language models in production environments RAG Implementation: Gained expertise in training models with custom datasets Docker & Cloud Deployment: Mastered containerization and multi-platform deployment Sentiment Analysis: Developed skills in natural language processing for review classification
Project Management Full-Stack Integration: Learned to seamlessly connect AI services with traditional web applications Performance Optimization: Understood the importance of balancing features with resource constraints User-Centric Design: Realized the value of automation in improving user experience
Industry Insights Hospitality Tech: Gained deep understanding of hotel management system requirements AI in Business: Learned how AI can solve real-world business problems effectively
What's next for hotel win win
Performance Enhancement: Upgrade to faster AI models or optimize current deployment Mobile Application: Develop native mobile apps for iOS and Android Advanced Analytics: Implement booking prediction and revenue optimization features Multi-language Support: Add internationalization for global users Predictive Analytics: Implement machine learning for dynamic pricing and demand forecasting Integration Ecosystem: Connect with popular hotel management systems and OTAs Voice Assistant: Add voice interaction capabilities to the chatbot
Hotel Win Win represents the future of hospitality technology - where artificial intelligence meets human hospitality to create exceptional experiences for everyone involved.
Log in or sign up for Devpost to join the conversation.