🌙 Ghibli Guardians — About the Project

✨ Inspiration

Personal safety is something many people think about every day, especially when walking alone at night, traveling, or navigating unfamiliar environments. We wanted to create something that felt more supportive than a typical safety app.

Our idea was inspired by the feeling of having a guardian companion: something calm, reassuring, and always available. Instead of focusing only on emergency situations, we wanted to build an experience that helps users stay calm, make better decisions, and feel emotionally supported when they need it most.

This vision led us to Ghibli Guardians, a safety companion that combines practical tools with comforting AI-driven interaction.

🛠️ How We Built It

We designed Ghibli Guardians as a full-stack application with clear separation between backend logic and frontend experience.

Backend

Built with Node.js + Express
Risk assessment engine based on real-world factors:
- time of day
- lighting conditions
- environment type
- whether the user is alone
Guardian contact system:
- users can add trusted contacts
- emergency messages are generated dynamically
Location tracking endpoint to store the latest coordinates
Emergency package generator:
- creates personalized messages for each guardian
- includes shareable Google Maps links
Rule-based chatbot assistant for safety guidance
AI voice integration using ElevenLabs API for character voice responses

Frontend

Interactive character-based UI
Chat interface connected to backend assistant
Voice playback for calming guidance
Emergency message generation with copy/share actions
Music and character experience designed to feel supportive rather than stressful

System Design Idea

The app works like this:

User Situation → Risk Analysis → Guidance + Voice Support → Emergency Message Preparation → Guardian Communication

⚡ Challenges We Faced

Building a real-time safety experience during a hackathon came with several challenges:

Integrating multiple APIs while keeping the system stable
Handling fallback logic when external AI services were unavailable
Designing backend endpoints that frontend teammates could integrate quickly
Balancing emotional design with practical safety functionality
Making sure the experience felt calm and human instead of alarming

One of the biggest challenges was adapting quickly when certain AI integrations failed. Instead of stopping progress, we redesigned parts of the system to work reliably with offline logic and fallback strategies. Needless to say, we were out of time, and our project needs improvement, but we'd like to use this space to write about what this project could have been and not what it was.

📚 What We Learned

Throughout this project we learned that:

Reliability is more important than complexity in safety-focused applications.
Good API design makes team collaboration much easier.
Emotional UX matters just as much as technical features.
Hackathons require fast iteration and flexible problem solving.
AI can enhance safety experiences when used as guidance rather than replacement.

Most importantly, we learned how to build something that combines technology with empathy.