Built an AI-powered restaurant agent that retrieves relevant menu information and generates responses using an
LLM-based workflow.
Implemented Retrieval-Augmented Generation (RAG) using vector embeddings and Supabase Vector Store to ground responses in restaurant menu data.
Integrated LLM APIs, prompt engineering, structured data, and automation workflows to connect retrieval, reasoning, and customer-facing responses.
Designed the workflow to handle Arabic restaurant menu queries including meals, additions, sauces, beverages, and combo items.
