Closer

Inspiration

We’re Beline, Louis, and Paul. While studying engineering in France, we volunteered with "Cheer Up", visiting children with cancer in hospitals and organizing activities with them. We saw how long stays separated them from school, friends, and everyday family life. We wanted to help them stay connected and support their emotional wellbeing between visits.

What it does

Closer connects hospitalized children with families, care teams, and volunteers through check-ins, shared questions, activities, and messages. AI suggests personalized conversation prompts and flags concerning responses for human review.

How we built it

We used Next.js and React for the interface, Supabase for authentication and data, and Mistral for AI features. We collaborated through GitHub and deployed with Vercel.

Challenges we ran into

Designing for four different roles, protecting children’s privacy, and making AI suggestions useful while keeping people responsible for decisions.

Accomplishments that we're proud of

Turning our volunteering experience into a working prototype that gives children, parents, volunteers, and care teams a shared space to connect.

What we learned

Small, consistent interactions matter. Useful AI can help someone notice a concern, ask a thoughtful question, or start a conversation.

What's next for Closer

Test with families and care teams, improve personalization, and evaluate whether our prompts help people feel more connected.

Built With

• claude
• claude-vision-model
• nextjs
• pytorch
• rl
• supabase
• typescript