Inspiration
Women often lack quick access to emergency assistance — delayed response times, no real-time tracking, unsafe commute routes, and little community awareness all add up to real danger in everyday situations. We wanted to build something that acts like an always-on safety companion, not just a panic button.
What it does
SafeWomen360 is an AI-powered women's safety ecosystem. Key features include:
- One-tap SOS and hands-free voice activation for emergencies
- Fake Call as an instant escape mechanism
- Live GPS location sharing with trusted contacts
- AI danger detection that listens for screaming or distress sounds
- Automatic police alerts to the nearest station
- Threat prediction using crime data, time, and location to flag risky areas before you enter them
- Deepfake voice calls that simulate a male voice to deter potential threats
- Auto recording of video/audio evidence the moment SOS is triggered, for legal use
- Community safety network that alerts other users within a 500m radius
How we built it
- Frontend: React Native and Flutter
- Backend: Node.js and Django
- AI/ML: TensorFlow and PyTorch for threat prediction and audio-based danger detection
- Cloud: AWS and Firebase for real-time data and storage
Challenges we ran into
Balancing real-time responsiveness (location tracking, distress detection) with battery and data efficiency, and training reliable threat-prediction and audio-detection models with limited data.
What we learned
How to combine multiple AI models (audio, prediction) into one responsive mobile experience, and how to design safety features that work even under stress or in low-connectivity situations.
What's next for SafeWomen360
Expanding the threat-prediction dataset, adding wearable device integration, and piloting the community alert network in a real neighborhood.
Log in or sign up for Devpost to join the conversation.