Care Circle

Inspiration

In the growing age of technology, it's easier for older adults or grandparents to fall behind. This often causes a disconnect between them and their grandchildren. We want to bridge this gap: create a simple platform that connects generations of family through technology, while protecting the elderly from any harm (i.e., scams) that comes with it.

What it does

Care Circle is a voice-based interface designed to help older adults navigate everyday tasks while staying connected with their families. Users can activate Care Circle through a regular phone call for everyday tasks (such as ordering groceries), while family members stay informed through a shared interface. Care Circle also includes a multi-layer fraud detection system that evaluates requests for signs of scams and pauses suspicious transactions for family verification. Beyond everyday assistance, the platform helps coordinate family calls and visits based on each person’s schedule, making technology a tool for strengthening connection.

How we built it

We built Care Circle as a TypeScript monorepo divided into separate voice, payments and fraud, family scheduling, and web services. The backend services use Fastify and PostgreSQL, while the frontend is built with Next.js and Tailwind CSS. We integrated tools for voice interaction, AI reasoning, payments, and video calling, including Twilio, Deepgram, Muse, LiveKit, and payment sandbox infrastructure. Shared API contracts allow each service to communicate while remaining independently testable.

Challenges we ran into

Our main bottlenecks in this project came from two things: getting the services to work together and making voice interactions reliable. Each part could work on its own, but threading a spoken request to an entire pipeline of pricing, fraud checks, payment, and family check-ins exposed subtleties that we overlooked. Voice was a completely different challenge. Dealing with people brings in a whole world of randomness and variability, making it more difficult to quantify tests and account for edge cases. Especially with a focus towards older adults, our voice detection and parsing had to be reliable and forgiving.

Accomplishments that we're proud of

We're most proud of integrating everything from voice assistance, everyday errands, and fraud protection all into one product/experience. There's still work to be done, but we built something that could actually help older adults stay independent while making it easier for their families to be there for them.

What we learned

We learned that designing for older adults isn't just about bigger buttons or fewer screens. At the end of the day, it was all about support. Small details, like how we ask for confirmation or explain a paused purchase, can make assistance feel reassuring rather than controlling. In short, it was less about what the app could do and more about how someone would actually feel whilst using it.

What's next for Care Circle

In the future, we want to complete the integration of Care Circle's services and test the entire experience using real voice, payment, and video-call infrastructure. We plan to fine-tune the fraud detection system, move components such as passkey verification and mock payments toward production-ready implementations, and continue expanding the AI assistant's functionalities. In the long term, we envision Care Circle as an all-encompassing everyday assistant for older individuals that keeps their families involved along the way.