Leak Alert

Project Story

Inspiration

Leak Alert was born from necessity in South Africa, a nation facing a R33 billion water crisis. With 40-50% of municipal water lost to undetected leaks, I realized we weren't just losing water; we were losing our national security. I was inspired to create a solution that empowers every citizen to become an active climate actor, turning a silent crisis into a battle we win together.

What it does

Leak Alert is an AI-powered water conservation ecosystem. It "closes the loop" on non-revenue water through a dual-track system:Citizen Engagement: Residents detect and report leaks in seconds via a live camera feed.AI Verification: A computer vision model provides real-time validation to filter false positives.Municipal Action: Verified alerts are routed to a live Command Center dashboard, directing maintenance teams to precise GPS coordinates.Marketplace: We connect users with a verified plumber marketplace for immediate B2C intervention.

How I built it

As a solo founder, I built the MVP and core architecture from the ground up. The system is powered by ASP.NET and Microsoft Azure, utilizing Advanced Data Analytics for predictive maintenance. I integrated a computer vision model to handle real-time leak validation and developed a scalable dashboard that visualizes sensor data into an intuitive map interface.

Challenges I faced

One of the biggest hurdles was managing the massive volume of data required for Non-Revenue Water analysis and ensuring the AI could accurately distinguish between a major burst and minor surface water. Balancing the technical development with the strategic pivot toward a B2B/B2C plumber marketplace required a significant shift in the business model to ensure long-term financial viability.

What I learned

I learned that technology alone isn't the answer—community power is. By turning residents into "active climate actors," we can achieve a 5:1 ROI on leak detection programs and avoid nearly 96 kg of $CO_2e$ per intervention.

Built With

asp.net
azure
computervision
cshtml
sql

Submitted to

The Climate Change-Makers Challenge: 2026

Created by

My Contribution
As the original visionary behind Leak Alert, my role focuses on the "Research" in R&D—bridging the gap between a high-level idea and a technically viable product. I am responsible for the conceptual architecture and the strategic roadmap that ensures our technology solves real-world water scarcity issues.

Key areas I led during this project:

Ideation & Conceptual Design: I identified the critical need for a decentralized, community-driven water management system. I conceptualized the core workflow: using AI and computer vision to empower citizens to report leaks in real-time, effectively turning every smartphone into a sensor for the city’s infrastructure.

R&D & Technical Logic: I defined the underlying logic for our detection algorithms. My research focused on how to differentiate active pipe bursts from environmental noise (like rain or shadows) to ensure our AI models provided high-accuracy data to municipal and private responders.

IP & Venture Readiness: I managed the intellectual property (IP) strategy during our TuksNovation and GIP funding applications. I focus on ensuring our R&D documentation meets the rigorous standards of legal due diligence and commercialization.

Tshepang95
Quincy Refentse

Updates

Quincy Refentse started this project — Apr 03, 2026 11:35 AM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.