AcousticLeakNet
2026CNN-based acoustic leak detection and localisation for water pipe networks, trained entirely on simulation and tested on three networks it never saw during training.
I'm a student at Delhi Public School Noida working on deep learning and cybersecurity. I've built an acoustic leak detector trained on 100M+ rows of simulated data, a malware analysis sandbox, and a daily quiz app for my school. I like taking an idea from paper to prototype to something that works in the real world.
★GENIUS Olympiad 2026 FinalistAcousticLeakNet was 1 of 27 projects selected from IndiaSelected work
CNN-based acoustic leak detection and localisation for water pipe networks, trained entirely on simulation and tested on three networks it never saw during training.
Malware analysis sandbox for Linux and Windows, built at GPCSSI. Runs a sample in a disposable QEMU VM, watches its behaviour and network traffic, scans it with YARA, and explains a 0–100 threat score. Qt desktop app and CLI.
Daily question-of-the-day quiz. Players sign in with Google and answer the day’s question, admins post new ones, and answers are graded automatically with manual override.
Digital assistant for Indian farmers covering soil health, crop recommendations, weather and market prices, with a multilingual interface. Built for the Inter DPS Digital Project Showcase.
Timeline
Gurugram Police Cyber Security Summer Internship (GPCSSI), certified.
Spotify 1980, a retro record store web app.
AcousticLeakNet, co-built with Aarav Gupta, was 1 of 27 projects selected from India.
Science stream: Physics, Chemistry, Math, CS.
Writing