AI

Hands-on AI projects I have built and run myself: agents that draft but wait for my approval, models that explain their scores, and the guardrails that make AI safe to use in real life.

Security

Satya-Vani: Catching a Cloned Voice While the Call Is Still On

Seven pluggable signals, one explainable 0–100 risk score that updates every half second: how I built live voice-spoof detection.

A phone showing an incoming WhatsApp message and a dashed AI draft reply waiting for Send, Edit or Ignore, next to the words 'Ask permission first' and an OFF to DRAFT to AUTO ladder.

I Built an AI That Replies to My WhatsApp, and Taught It to Ask Permission First

A self-hosted agent that drafts WhatsApp replies in my voice. Every chat starts OFF, trust is earned per chat, and the bot always says when it's the bot.