SafeHer AI

Inspiration

Personal safety, especially for women, remains a critical challenge. In emergency situations, victims often do not have the time or ability to unlock a phone, open an app, and manually press an SOS button. This inspired the creation of SafeHer AI—a proactive, intelligent safety assistant that bypasses traditional barriers to automate emergency communication when every second counts.

What it does

SafeHer AI is a smart personal safety web application that actively monitors for distress signals and triggers immediate emergency actions.

Multi-Modal Triggers: Activates an SOS via Voice Commands, Audio Distress Detection (screaming/panic sounds), or a physical Phone Shake.
Instant Emergency Actions: Once a trigger is detected, it immediately activates a Loud Siren Alarm to deter attackers, shares a Live Location Map, and sends automated SOS Messages.
Automated Emergency Communication: It initiates an Auto-Call to pre-saved emergency contacts to ensure immediate help is on the way. ## How we built it Building an application with deeply integrated device features (like hardware sensors, microphone access, and communication APIs) traditionally requires complex full-stack mobile development.

Using MeDo, we built this entire platform using natural language prompts:

Conversational Scaffolding: We described the core logic to MeDo—how to handle background microphone listening for audio distress and sensor mapping for phone shakes.
Component Generation: MeDo generated a seamless user interface integrating a responsive live map component, local storage for emergency contacts, and Web Audio API scripts for the high-frequency siren.
Plugin Integration: Leveraged MeDo's ecosystem to efficiently mock web-based geolocation sharing and emergency communication triggers. ## Challenges we ran into The biggest challenge was ensuring the background triggers (voice and shake detection) operate with high accuracy without causing false positives. Fine-tuning the threshold for "phone shaking" and optimizing continuous audio detection in a web-based architecture required precise prompt iterations with MeDo's multi-turn chat to clean up the event listener loops. ## Accomplishments that we're proud of We successfully transformed a complex mobile safety architecture into a rapidly deployable, lightweight web application using MeDo. The seamless transition from audio/sensor input to automated emergency communication works flawlessly in real-time. ## What we learned We learned the power of intent-driven development. By letting MeDo handle the heavy lifting of UI layout, API boilerplate, and core JavaScript event listeners, we were able to focus entirely on perfecting the user experience and the critical safety workflows.