Buildinators - AI à La Carte - Redis Mini Hack Dubai — Oct 28, 2025
AI Tinkerers - Dubai
Hackathon Showcase

Buildinators - AI à La Carte

Team consisting of Fatima Group AI/MLOps Engineer Muhammad Haris Salman and Harmen Labs Lead Data Analyst Aimen Imran — Python, SQL, Firebase, Redis, Looker.

2 members

This project tackles the critical need for instantaneous, error-free customer service in the UAE’s rapidly expanding hospitality sector. The core problem is simple: high-volume restaurants and hotels need to answer complex, multilingual guest questions (e.g., “Do you have any vegan items?”) instantly, without incurring huge LLM API costs or relying on slow, old databases.

Our solution, the AI a La Carte, uses Redis as a real-time data layer and memory manager to deliver sub-second response times and dramatically reduce operational expenses.

We are solving for latency and language barriers—two major friction points in a global hub like Dubai—by enabling staff and automated systems to provide accurate menu information across multiple languages instantly.

Google Gemini API: Used gemini-2.5-flash-preview-09-2025 for all AI logic (Dynamic Query Classification JavaScript/HTML/Tailwind CSS: Responsive frontend interface. Multimodal Analysis Node.js/Express: Simple Redis Redis Set (SET): For Real-time Inventory Management (inventory:oos). Redis Sorted Set (ZSET): For Popularity Ranking (dish:ratings). Redis: The core engine for all data indexing Vector Index (RediSearch Module): For the Semantic Cache and Menu RAG. and RAG Answer Generation). and real-time operations. Specifically: caching fast backend for handling API routing and integration logic.