Teen Coder Turns Pothole Nightmare Into AI-Powered Civic Tool

When a motorcycle accident on a New Delhi pothole sent his parents to the hospital, 15-year-old Parth could’ve done what most of us do: complain on social media and move on. Instead, he built an app.

Project Sadak is the kind of solution that makes you wonder why it didn’t exist sooner. The platform lets citizens photograph potholes, use GPS to pinpoint their location, rate severity, and—here’s the kicker—automatically draft and send a formal complaint to local authorities. No hunting for email addresses, no bureaucratic runaround. Just snap, submit, and let AI handle the paperwork.

The technical backbone is cleverly layered. AI validates that a submitted image actually shows a pothole, not some random road texture. Human reviewers then grade each one as severe, medium, or benign, ensuring the worst hazards flash red on the map rather than getting lost in a sea of minor complaints. Once verified, users digitally sign an auto-generated email and send it straight to the relevant New Delhi officials.

Since launch, the platform has logged 360 pothole reports, with 11 repairs completed—though Parth admits he’s personally funded most of those fixes through contractors connected to his father’s construction business. That detail reveals the frustration at the heart of this story: a 15-year-old felt compelled to literally fix the roads himself because bureaucratic dysfunction and diffusion of responsibility made waiting for government action futile. There’s a pothole reported even in Bengaluru that got repaired, despite AI struggling to locate the right contacts.

Looking ahead, Parth envisions full automation from photo to email dispatch, and a WhatsApp chatbot to eliminate the need for app downloads altogether. He’s also aware of a current gap: the platform lacks a follow-up system to track whether reported potholes actually get fixed. That’s the next frontier—not just reporting the problem, but proving the system works.