Contact person: Women Safety Ecosystem

Phone:Show

Email:Show

Location: Mumbai, India

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"Buil an end-to-end personal safety platform that unites three equal-priority components—an AI-driven CCTV system, a feature-rich mobile application, and a biometric watch—into one seamless network.

AI CCTV
The camera backend must automatically detect people in the frame, classify gender, recognise potential threat scenarios in real time, and immediately push alerts to the rest of the system. Accuracy, low latency, and clear event logging are essential.

Mobile App
A single shake should trigger SOS, fire off a siren, and broadcast the alert to nearby users through a community network. The app also needs:
• Safe-route navigation with live rerouting when risks are detected
• Audio- and video-record on command or automatically during an incident
• Fake-call generator for de-escalation
• Offline functionality that queues data until connectivity returns
• AI self-defence chatbot and one-tap emergency dialling



Core Deliverables
1. End-to-end prototype or MVP that links all three devices, demonstrates real-time alert flow, and stores events securely.
2. Source code, model weights, and build/installation instructions.
3. Short demonstration video or live walkthrough validating people detection, threat classification, SOS triggers, and cross-device synchronisation.

Acceptance Criteria
• Person detection F1 ≥ 0.9 under typical indoor lighting.
• Alert latency ≤ 2 seconds from event to mobile notification.
• App continues to function (record, siren, SOS queue) with no data for at least 30 minutes offline.


Feel free to propose the most suitable stack—e.g., Python or C++ for edge inference, Flutter or React Native for the app, —as long as the final build meets the performance targets above." (client-provided description)