Building SaveYah: A Community-Powered Emergency Response Platform at the MeDo Hackathon

Hackathons often challenge developers to build fast. But sometimes, they also challenge us to build with empathy.

During the MeDo Hackathon, I built SaveYah — a real-time emergency response platform designed to help communities coordinate faster during emergencies while also providing vulnerable individuals with safer and more discreet ways to seek help.

The inspiration behind SaveYah came from observing how emergencies are often handled in many communities: fragmented WhatsApp messages, delayed communication, limited visibility, and slow response coordination. In critical moments, information usually reaches social media before it reaches the people who can actually help.

I wanted to explore how technology could help bridge that gap.

SaveYah allows users to:

• Report emergencies in real time
• View live incidents on interactive maps
• Discover nearby emergency resources
• Verify incidents through community participation
• Receive realtime alerts and updates
• Trigger SOS emergency workflows discreetly

One of the most impactful features I implemented was a hidden GBV emergency panic system.

With rising cases of gender-based violence in Kenya, I wanted to think beyond traditional panic buttons, which can sometimes place victims at greater risk if discovered. In SaveYah, the panic trigger is intentionally hidden behind a normal weather widget. When tapped three times quickly, the platform silently sends the victim’s live location and a short emergency audio recording to trusted contacts through WhatsApp — without exposing the alert publicly on the platform feed.

This feature pushed me to think deeply about safety-centered design, discretion, and accessibility in emergency technology.

From a technical perspective, SaveYah was built using a modern JAMstack architecture powered by:

• React + TypeScript
• Vite + Tailwind CSS
• Supabase Realtime + PostgreSQL
• Supabase Edge Functions
• Leaflet.js for mapping
• Gemini 2.5 Flash for AI-assisted incident classification
• Twilio API for WhatsApp emergency communication

Participating in the MeDo Hackathon also gave me the opportunity to explore AI-assisted development workflows in a practical way. Using MeDo accelerated rapid prototyping, UI iteration, and feature experimentation, allowing me to focus more on problem-solving, architecture, and user experience during the limited hackathon timeline.

One of the biggest lessons from building SaveYah is that emergency systems are not only technical challenges — they are human-centered challenges. Designing for moments of fear, urgency, and vulnerability requires empathy just as much as engineering.

SaveYah started as a hackathon project, but I believe the idea has potential far beyond the competition. My long-term vision is to evolve the platform into a scalable community emergency coordination system with broader integrations for responders, NGOs, and local support networks.

Technology becomes most meaningful when it helps people feel safer, more connected, and more supported in moments that matter most.

SaveYah #BuiltWithMeDo #Hackathon #TechForGood #EmergencyResponse #CivicTech #ReactJS #Supabase #AI #Kenya #GBV