I'm a 2nd year ECE student at KIT, Tamil Nadu.
I just shipped VoiceID — a free tool that detects
whether a voice is human or AI-generated.
The Problem
Deepfake audio is becoming a real threat. Most detection
tools only support English. Nobody was building for
Indian languages.
What I Built
VoiceID analyses 88 acoustic features from any MP3 or
WAV file and classifies the voice as Human or AI with
a confidence score.
Supports: English, Tamil, Hindi, Malayalam, Telugu.
Live demo: https://mdeepaksai.github.io/human-or-AI/
Tech Stack
- FastAPI backend deployed on Railway
- Vanilla JS frontend on GitHub Pages
- Supabase for live visitor tracking
What I Learned
Shipping a real product is completely different from
building a college project. You deal with CORS issues,
file size limits, audio format compatibility, and
latency that you never think about in a demo.
The most valuable thing wasn't the ML model.
It was deploying something real people actually use.
Try It
Free. No login. No signup.
https://mdeepaksai.github.io/human-or-AI/
Top comments (0)