Have you ever ordered a dish at a restaurant, just to find out it's much smaller, larger, or less appealing than what you expected? Menus gives us limited representations of the food on the menu - we wanted to find a way to see our food before we order it. This could not only improve the dining experience at restaurants, but also in the growing online food delivery market.
We created a meal recommendation app that finds the perfect meal for you through an AI chat assistant, and displays available dishes right on your countertop through augmented reality. When you're at a restaurant, you can see a preview of their dishes in their real scale and appearance in AR. When you're at home, our smart ordering chatbot finds the perfect meal that matches your tastes, and shows you your recommendations in full AR.
We built our iOS app using Swift 4, and Arkit. We scanned actual sushi, burgers, hotdogs, side dishes and dessert at Hacktech using the Qlone app, and processed them in Blender to prepare them for AR. We built our AI chatbot on the Azure cloud, with Node.js.
Most of us had to learn Swift and XCode from close to scratch, so we had to muster through learning new syntax, IDEs, and app development patterns. We also needed to experiment to get high quality scans with the Qlone app, which including several hours of turning sushi on 3' by 3' paper scanning mats we taped together.
We're proud of creating a polished Swift app, scanning high fidelity models of a dozen foods, and creating an intelligent chatbot. We're also really proud of branching out of our silos and sharing our expertise with each other as a team.
We know now that XCode is loaded with traps and pitfalls we need to learn to avoid.
We hoping to increase our menu repertoire, simplify the 3D scanning process, and improve our chatbot's recommendation abilities. We're excited by the vanishing barrier between the physical and the digital!