Auri is designed as a real-world AI assistant that connects fragmented data across platforms and turns it into actionable intelligence. It integrates systems like calendars, email, and contextual inputs to build a structured understanding of the user’s day. A priority engine and time-based filtering ensure that only critical information is surfaced, reducing cognitive load. Beyond information delivery, Auri supports real actions such as preparation, scheduling, and communication, demonstrating how AI can move from passive tools to active decision-support systems.

This project is an emergency medical collaboration app, stemming from a real-life family medical tragedy. The application focuses on the entire process from "onset of illness—emergency treatment—transportation—treatment—rehabilitation." It integrates family emergency guidance, wearable device data, and connects with ambulance and hospital systems to achieve more efficient medical response and tiered healthcare. Clear guidance on hospital capabilities, it helps to appropriately distribute patients, alleviate the overload on top-tier hospitals in China, and ensure that more people receive timely and effective medical treatment.

MediMate reimagines medication management as a connected system rather than a single device.
It addresses a critical real-world problem: elderly users often forget, skip, or accidentally repeat medication, while caregivers face ongoing stress managing daily routines.

The solution integrates a high-capacity smart dispenser with a multi-user mobile application.
Through synchronized light, voice, and visual feedback, the system provides clear, real-time guidance for each user. It also supports family-level coordination, enabling shared monitoring, remote notifications, and medication tracking.

Unlike traditional pillboxes, MediMate emphasizes system thinking—bridging hardware interaction, digital interfaces, and behavioral design.
The project focuses on reliability, clarity under pressure, and reducing cognitive load, especially for aging users.