ELDERGUARD AI

DASHBOARD WIH MULTI LANGUAGE SUPPORT
LOGIN PAGE
EMOTION DETECTOR
DAILY ALARM SCHEDULE
VOICEBO-ACT AS A COMPANION
PROGRESS
CAREGIVER DASHBOARD

Inspiration

As technology advances, one group that can easily be left behind is the aging population. Many elderly people live alone or require regular assistance with medication, emergencies, health monitoring, and everyday activities. At the same time, caregivers cannot be available every minute of the day.

This inspired us to build ElderGuard AI — an intelligent digital guardian designed to help seniors live more safely and independently while keeping their caregivers connected.

Our goal was simple: use AI not just to automate tasks, but to provide meaningful care when it matters most.

What We Built

ElderGuard AI is an AI-powered elderly care and safety platform that combines Generative AI, Agentic AI, Voice AI, and real-time monitoring into one system.

Key features include:

🧓 AI Companion – Mitra: A conversational voice-enabled companion that allows seniors to interact naturally and receive assistance.
❤️ Health Monitoring: Monitors health-related information and provides a simple overview of the user's well-being.
🚨 Fall Detection & Emergency Alerts: Detects potential falls and triggers emergency workflows to notify caregivers.
💊 Medication Reminders: Reminds seniors to take their medications and allows them to mark them as taken.
🆘 SOS System: Provides a quick emergency mechanism for contacting designated caregivers.
👨‍👩‍👧 Care Circle: Keeps trusted family members or caregivers connected to important alerts and activities.
🧠 Digital Health Twin: Maintains an AI-accessible representation of the user's health history and relevant information to provide more personalized assistance.
📊 Activity & Health Overview: Presents recent activities, alerts, reminders, and a simplified health score through an easy-to-use dashboard.

How We Built It

We designed ElderGuard AI as a modular system where different components work together as an intelligent care ecosystem.

The frontend was built using React, TypeScript, Vite, and Tailwind CSS, with a focus on creating a simple interface that is comfortable for elderly users.

For the backend, we developed a Java 21 REST API to handle users, emergency contacts, reminders, alerts, and communication between the application components.

We also experimented with agent-based workflows through SNS Workbench, allowing different AI agents and services to work together for tasks such as monitoring, alerts, and assistance.

The system follows a workflow where:

User/Health Data → Monitoring → AI/Agent Processing → Risk Detection → Alert Generation → Caregiver Notification

This architecture allows ElderGuard AI to move beyond being a simple chatbot and act as an active digital care assistant.

Challenges We Faced

One of the biggest challenges was designing a system that could be powerful enough to provide intelligent assistance while remaining simple enough for elderly users.

We also faced several technical challenges while integrating different components, including:

Connecting the frontend with our Java backend.
Designing reliable emergency and fall-detection workflows.
Integrating AI agents through SNS Workbench.
Handling webhook communication between services.
Testing real-time alert scenarios.
Managing backend data and user information.
Making the interface accessible and easy to understand.
Exposing and testing our local backend during development using ngrok.

Building the project within a limited hackathon timeframe also forced us to prioritize the most important features and rapidly iterate on our architecture.

What We Learned

This project taught us that building an AI application is not only about choosing a powerful model. System design, user experience, reliability, and responsible automation are equally important.

We learned how to:

Design an AI-powered healthcare-oriented application.
Build and connect REST APIs.
Work with agentic AI workflows.
Integrate frontend, backend, and AI services.
Design event-driven emergency workflows.
Think about AI from a real user's perspective rather than only from a technical perspective.
Rapidly prototype and iterate under hackathon constraints.

Most importantly, we learned that technology can be designed to support human care rather than replace it.

The Vision

ElderGuard AI is more than an emergency alert system. Our vision is to create an intelligent companion that helps seniors maintain their safety, independence, and connection with their loved ones.

We believe aging should not mean losing independence.