Inspiration

The inspiration for Stay Close came from a deeply personal place. Watching elderly family members live alone while their children worry from afar is a reality millions face daily. I realized that technology could bridge this emotional distance, not through complex interfaces, but through simple, meaningful daily connections. The question became: how can we make wellness monitoring feel like love, not surveillance?

What it does

Stay Close creates gentle touchpoints between elderly individuals and their families through two intuitive interfaces:

For Elderly Users:

  • Daily health check-ins with photo sharing to show they're doing well
  • AI-powered exercise recommendations based on specific pain areas and preferred locations
  • Family event calendar with reminders for birthdays, appointments, and gatherings
  • Emergency assistance button for immediate family notification

For Family Members:

  • Daily wellness dashboard showing check-in status and health trends
  • Synchronized family events to coordinate care and celebrations
  • Peace of mind monitoring without being intrusive

How I built it

I built Stay Close primarily using Bolt.new, leveraging its rapid development capabilities to create a responsive, user-friendly application. The tech stack includes:

  • Frontend: React with clean, accessible UI design optimized for elderly users
  • Backend: Supabase for real-time data synchronization between family members
  • AI Integration: Custom exercise recommendation engine based on user input with Gemini Flash 2.0 model
  • Deployment: Netlify for reliable, fast global access

Challenges I ran into

  • Accessibility design: Creating interfaces that work for elderly users
  • Data synchronization: Ensuring real-time updates between family members and Elderly
  • AI exercise generation: Building meaningful, safe exercise recommendations that adapt to individual needs and limitations
  • Emotional design: Balancing monitoring capabilities with maintaining dignity and independence for elderly users

Accomplishments that I am proud of

  • Intuitive role-based interface that makes technology approachable for all generations
  • Seamless family connection without feeling invasive or overwhelming
  • Responsive design that works beautifully on both smartphones and tablets
  • Real-world applicability - this isn't just a hackathon project, it's a solution families actually need
  • Emotional resonance - turning daily wellness into moments of connection

What I learned

Building Stay Close taught us valuable lessons about AI-assisted development and strategic planning:

  • Plan first, prompt second - Define your product flow and stick to it before writing any prompts, otherwise AI will lead you in circles without clear direction
  • Database-first approach - Build and validate your database schema early using tools like Claude MCP, then regularly verify it stays correct throughout development
  • Feature before UI - Focus on building core functionality and testing it thoroughly before polishing the UI - this prevents getting distracted by aesthetics over substance
  • Break down complex requests - Smaller, specific prompts yield better results than large, multi-part requests that can confuse AI assistance
  • AI over-assistance awareness - Even with clear instructions like "CHANGE ONLY THE WORDING," AI tools sometimes add unintended features - constant vigilance and the revert function in Bolt.new are essential
  • Workflow optimization - Using consistent conversation contexts (like staying within one Claude project) saves significant time when iterating on prompts and maintaining project continuity These insights transformed our development approach from reactive prompt-fixing to strategic AI-assisted building.

What's next for Stay Close - Family Care App for Elderly

  • Medication reminders with family notification systems
  • Video calling integration for face-to-face check-ins
  • Community features connecting elderly users with similar interests and local activities
  • Multi-language support to serve diverse family structures
  • Telehealth integration for seamless medical consultations

Built With

Share this project:

Updates