SafeSangkum

Situation

Ask someone in Cambodia what they would do in an emergency and the answers come back the same way: no emergency contacts saved, no idea which number to call, and a real fear of being unable to request rescue at the moment it is needed. The gap is not the emergency services. It is the thirty seconds between something going wrong and reaching anyone at all.

The people who feel this most are young adults aged 18 to 29, and especially late-hour and night-shift workers and students travelling the roads late in the evening. They are comfortable with location-sharing technology. Nobody had pointed it at their safety.

Task

I am the founder of SafeSangkum, built with a team of four through Techo Startup Center's Digital Platform Accelerator Cohort 2. As founder and frontend developer, I owned the product: the interface, the emergency flows, and the decisions about what a person should have to do when they are least able to do anything.

The brief we set ourselves was that requesting help should be one action, not a decision tree.

Action

The design runs on one principle: in an emergency, the user should not have to choose between channels, and should barely have to aim.

• Shake detection. Shake the device three times and it shares your location with the nearest police station after a ten second countdown, with the option to notify connected friends that you are in danger. The countdown is the safety valve, and the button layout is customisable so the priority order is the user's own.
• A trusted circle, built beforehand. Friends connect by QR code, so the circle exists before it is needed rather than being assembled in a crisis.
• Organised emergency contacts. Fire engine, police, ambulance, water rescue, mental health (TPO), and public works on one screen, each an icon, no menu.
• A community feed. Requests and reports posted anonymously: thefts near a market, a blood donation needed at a named hospital, a breakdown on a road. Help that does not need a siren, from people already nearby.

We built it in Flutter and Dart on AWS, with end-to-end encryption and SHA256 hashing of public usernames, because an app that knows where you are and who your family is has to be trustworthy before it is useful.

Against the alternatives, the gap was clear. Find My offers organised contacts but cannot notify friends of danger or request emergency help; Facebook is where Cambodians already report problems, but not anonymously and not privately.

Result

SafeSangkum presented at Demo Day in Phnom Penh on 22 June 2023, closing the accelerator with a certificate of appreciation for the team, on a business model built around data insights for NGOs and verified in-app advertising rather than charging people for safety.

The survey responses shaped the roadmap more than the pitch did. People asked for Khmer language support, for the app to work without an internet connection, and for the emergency hotline numbers to be advertised more widely. The last one is the finding I keep coming back to: part of the problem was never technical, and a product that makes existing services easier to find is doing real work before it ships a single new feature.