Made by Cedric Garand, Pooria Roy, Dolev Klein and Jaskirat Kaur.
Wellness Cowboy is an AI-powered stress management coach that runs on an Amazon Echo Dot, designed to help users stay productive and healthy while they work. Using computer vision, it continuously monitors a user’s facial expressions to determine a stress index. When stress levels get too high, the Echo Dot activates "Cowboy Coach" mode, taking over the user's screen and guiding them through relaxation exercises with voice instructions and on-screen tutorials—all with a classic cowboy twist. ðŸ¤
To build this, we developed a custom Alexa skill that processes stress data, triggers voice responses, and integrates with GPT-4o to generate the right relaxation exercises. The backend uses OpenCV and MediaPipe for facial analysis, FastAPI for communication, and OpenAI’s models for the exercice choice. The frontend was initially designed in Figma and later implemented using HTML, CSS, and JavaScript. Along the way, we faced challenges, from unreliable Echo Dot pairing to navigating Alexa’s wild-west documentation, but we persevered. 🤠YEEHAW !
Made by Cedric Garand, Anas Bourfia and Yassir Maknoun.
SerenED is an innovative solution designed to improve the patient and companion experience in emergency departments (ED) by transforming the traditionally stressful waiting time into an engaging and supportive experience. The platform features a page that provides real-time updates on wait times and patient status, a minigame to occupy children, an interactive social system for connecting patients in similar situations, and a breathing exercice to calm patients. Built with React and real-time updates via Socket.IO, SerenED offers a seamless, responsive interface that fosters connection, reduces anxiety, and enhances emotional well-being while waiting. The design is fully optimized for both PC and mobile devices, ensuring accessibility and functionality across different screen sizes.
Made by Cedric Garand, Sahij Gill, Nawaf Ahmed and Krish Shah.
BlockedIn is an AI-assisted Tetris game designed to balance human involvement and automation, inspired by real-world challenges in Air Traffic Control. The AI provides two placement options for each piece—one optimal and one less optimal—requiring the player to make the final decision, ensuring accountability and engagement. Using Groq-powered AI, OpenCV eye tracking, and emotional analysis, the game monitors player focus in real time. If focus drops, the game enters "Focus Mode," temporarily removing hints to encourage independent decision-making. Built with HTML, CSS, JavaScript, and Python, BlockedIn demonstrates how AI can enhance cognitive engagement while keeping human players actively involved.
Made by Cedric Garand, Sahij Gill, Nawaf Ahmed and Moveen Binuwara.
Avalanche is an AI-integrated journaling platform designed to help users make gradual, meaningful changes to their routines. By analyzing daily journal entries and key personal details—such as goals, hobbies, and interests—Avalanche provides personalized, small-scale advice that helps users improve by 1% each day. The platform features text-to-speech input for accessibility, a chat history for tracking progress, and an intuitive design for ease of use. Built using Groq-powered AI, Avalanche ensures that recommendations are relevant and actionable. Future plans include enhanced AI customization, deeper user insights, and mobile app integration.