CareLoop: AI-Powered Elderly Care Assistant

💡 Inspiration

Caring for elderly family members involves managing medications, appointments, daily routines, and communication between caregivers. These responsibilities can become overwhelming, especially when family members cannot always be physically present. We wanted to explore how AI and voice technology could make everyday caregiving simpler and help families stay informed.

This inspired us to build CareLoop, a voice-first AI assistant designed to help caregivers organize elderly care through natural conversations.

🚀 What It Does

CareLoop brings essential care-management tasks into one place. It includes:

• Medication tracking: Check medication schedules and record whether medications have been taken.
• Appointment management: View upcoming appointments and prepare new bookings with confirmation.
• Daily care summaries: Generate summaries of important care information.
• Family check-ins: Request check-ins and prepare alerts for family members.
• Conversational assistance: Interact with the assistant through a browser-based voice interface or text chat.
• AI integration: Use Amazon Bedrock for AI-powered summaries when properly configured.

The application also exposes care-management capabilities through the Model Context Protocol (MCP), allowing compatible AI clients to interact with its tools.

🛠️ How We Built It

We developed CareLoop using TypeScript, Node.js, Express.js, HTML, CSS, and JavaScript.

The backend provides REST API endpoints for the web interface and an MCP endpoint for compatible AI clients. SQLite stores care records, while synthetic sample data helps demonstrate the workflows. The frontend uses browser speech capabilities for voice input and spoken responses. Amazon Bedrock integration supports AI-generated summaries when the required AWS configuration is available.

We also implemented confirmation-based workflows for sensitive actions, created integration tests, and deployed the application through GitHub and Render.

Technology stack:

• Languages: TypeScript, JavaScript, HTML, CSS
• Backend: Node.js and Express.js
• Database: SQLite
• AI service: Amazon Bedrock
• AI tool integration: Model Context Protocol (MCP)
• Voice interface: Web Speech APIs
• Hosting: Render
• Version control: Git and GitHub

📚 What We Learned

Building CareLoop helped us understand how to connect AI systems with practical application tools, design APIs, manage structured data, and create conversational interfaces. We learned how MCP can expose application capabilities to compatible AI clients and how confirmation steps can help prevent unintended actions.

We also gained experience with environment variables, deployment configuration, testing, and preparing a full-stack application for public access.

🧩 Challenges We Faced

One major challenge was connecting the different parts of the application into a working system. We had to coordinate the frontend, backend, SQLite storage, AI integration, and MCP communication.

Deploying the project introduced additional challenges involving GitHub authentication, environment configuration, build commands, and running the server on a cloud platform. We addressed these by debugging the deployment process and successfully publishing the application on Render.

Another important consideration was reliability and privacy. Care-related information must be protected, AI-generated summaries should not replace professional medical advice, and external notifications must not be presented as delivered unless delivery is actually implemented.

🌟 What's Next

We would like to improve CareLoop with stronger authentication, persistent cloud storage, more reliable family notifications, and deeper integration with compatible voice-assistant platforms. Our goal is to make

Built With

• amazon
• api
• bedrock
• context
• css
• express.js
• html
• javascript
• mcp
• model
• node.js
• protocol
• render
• rest
• speech
• sqlite
• typescript
• web