Aurago

Discovering food should feel effortless, intuitive, and personal — yet today’s food delivery platforms still rely on old search technology that treats meals like plain text. Users have to guess keywords, scroll endlessly, and hope the right dish appears. Restaurants struggle to make their menus “search-friendly,” even though what truly matters is how the food tastes, feels, and satisfies a craving. This project introduces a completely new way to find food: an AI-powered craving engine that understands the meaning and intent behind what the user wants. Instead of relying on keyword matching, the system interprets natural language, reasons about context, and retrieves dishes that best satisfy the user’s request. Whether someone says “I want something spicy but not too heavy,” “a comforting soup like my grandmother used to make,” or “a protein-packed lunch that isn’t boring,” the app goes beyond the text and grasps the underlying craving. It works like a chef who knows flavor, mood, and texture — but instead of cooking, it curates the perfect meal from restaurants near the user. Traditional platforms rely heavily on search engines like Elasticsearch or AWS OpenSearch. While powerful for structured text retrieval, they do not understand meaning. They cannot differentiate between cravings, moods, or nuanced descriptions of food. As a result, users must filter manually, adjust keywords, or browse endlessly. This system flips the model entirely. It begins with understanding, extracting the user’s intent with a semantic model that interprets emotions, context, cuisine preferences, dietary requirements, and descriptive attributes of food. Once the craving is fully understood, the engine intelligently retrieves the closest matching dishes. The implications are transformative: Users communicate naturally, without needing keywords or filters. Restaurants are encouraged to describe dishes more intentionally, focusing on flavor, experience, and uniqueness rather than SEO-style text. Food discovery becomes dynamic and personalized, similar to how ChatGPT revolutionized information retrieval by understanding context rather than scanning for keywords. In many ways, the system acts as the “ChatGPT of food apps.” While existing delivery platforms operate like traditional search engines — hoping users will sift through long lists — this app thinks about the request, interprets nuance, and returns meaningful results. The result is a faster, more accurate, and more human way to browse food. Users feel understood. Restaurants feel represented. And the entire search experience becomes more intuitive, intentional, and enjoyable. This prototype is only the beginning. Alongside this intelligent ordering experience, the project also includes ongoing development of a companion restaurant app and a driver application. Together, they form the foundation of a next-generation food delivery ecosystem — one built around AI reasoning, seamless workflows, and an elevated user experience. As the system evolves, it has the potential to reshape how people discover food, how menus are structured, and how restaurants present their offerings. With continued refinement, guidance, and real-world testing, this project could become a standout innovation in a rapidly growing market. I would be excited to continue building and improving this project with your guidance even beyond the course — because the vision extends far beyond a class assignment. It represents a new, more intelligent future for food discovery.