-
Logo
-
Home Screen (Dashboard)
-
Page 1
-
Page 2
Inspiration
The idea for AstroCare came from a personal experience that highlighted how critical early detection can be in saving lives. After losing my kitten, Astro, to feline panleukopenia, I realized that much of the tragedy stemmed from not catching the warning signs early enough. That moment made me wonder: what if technology could help detect subtle signals of distress sooner — in both pets and people? This project grew from that question. I wanted to build something that not only leverages my passion for AI, medicine, and engineering but also carries a very real purpose: creating accessible tools that monitor health, alert families, and connect patients with providers before emergencies escalate. AstroCare is my way of turning a painful experience into motivation to make a difference in how we care for those we love.
What it does
AstroCare is an AI-powered interactive health companion designed to detect, connect, and personalize care. It uses real-time monitoring through camera and microphone input to flag potential emergencies, such as irregular breathing, falls, or slurred speech. When an anomaly is detected, the system raises an alert on a live dashboard, allows users to notify caregivers or healthcare providers instantly, and provides personalized risk scores, health trends, and care notes for context. A built-in assistant chatbot also explains alerts and guides next steps in simple language, making the experience both interactive and actionable.
How we built it
I built AstroCare as a React and Vite web app, styled with Tailwind, and powered by lightweight AI/ML prototypes using microphone signal analysis and event simulation. The prototype includes a live monitoring dashboard with charts built using Recharts, webcam and microphone integration for real-time input, and event simulation buttons to demonstrate detection of falls, irregular breathing, and slurred speech. I also designed a rule-based chatbot to provide accessible guidance during alerts. Privacy was an important design principle, so the prototype ensures that all data remains local to the device during demonstrations.
Challenges we ran into
One of the biggest challenges was getting real-time audio and video processing stable in the browser within the short timeframe of the hackathon. I also had to carefully balance technical complexity with accessibility so the demo could remain beginner-friendly while still making an impact. The time pressure of designing, coding, and polishing a working prototype in just a couple of days was significant.
Accomplishments that we're proud of
I am proud that I was able to deliver a fully interactive prototype with real-time monitoring and visualization in such a short period of time. More than just coding, this project represents my ability to design a tool that addresses a real-world problem in healthcare and personal safety. I am also proud of creating a clean, intuitive user interface that makes the experience clear and approachable, and most importantly, of transforming a personal loss into something that could help others prevent emergencies.
What we learned
Through this project, I learned how to integrate multimedia inputs like cameras and microphones into a web application and how to use data visualization to make health information more accessible. I also realized the importance of simplicity in prototypes: focusing on a functional and clear demonstration communicates the concept far better than over-complicating features under time constraints. Perhaps the most valuable lesson was that storytelling and personal impact are just as important as technical skill in building solutions that matter.
What's next for AstroCare
The next steps for AstroCare include replacing the demo heuristics with real TensorFlow.js models for detecting events such as coughing, sleep apnea, and falls through pose estimation. I also want to add secure authentication and support for exporting health data in standardized formats such as FHIR to make the system interoperable with existing healthcare solutions. Beyond that, I hope to expand personalization by training machine learning models to provide more accurate risk predictions over time, and eventually create a mobile-first version so the tool can be accessed more widely. Inspired by Astro, I am also interested in exploring potential applications in veterinary medicine to detect health issues in pets, completing the circle of where this journey began.
Log in or sign up for Devpost to join the conversation.