> whoami
> building_tonelens.exe
> loading_story...

THE STORY BEHIND TONELENS

Built in 7 days. I just kept going until it worked.
THE PROBLEM
/* * People miss 70% of what's really being said. * A nervous laugh. A confident tone hiding desperation. * Crossed arms. Avoiding eye contact. * * Google Translate gives you words. * ToneLens gives you truth. */
THE BUILDER
$ cat builder.json
{
  "name": "Mohan Prasath",
  "age": 18,
  "location": "Chennai, India 🇮🇳",
  "education": "B.Tech CSE (AI & ML)",
  "interests": [
    "edge_ai",
    "web_development",
    "entrepreneurship",
    "hackathons"
  ],
  "status": "building_the_future"
}
BUILD TIMELINE
[DAY 1]
2026-03-09 - Initial commit, then the first end-to-end ToneLens app landed with backend, frontend, and deploy setup.
[DAY 2]
2026-03-09 - Meeting consent UI, screen-share audio, presentation mode, and the history/emotion UI were added.
[DAY 3]
2026-03-09 - Strict response formatting, the Vertex AI reformat step, and the first README/demo architecture pass shipped.
[DAY 4]
2026-03-10 - Agent tools and bridge were refactored, lazy Vertex AI init was added, and negotiate mode plus extra pages landed.
[DAY 5]
2026-03-13 - Frontend pages were cleaned up and the landing layout was tightened.
[DAY 6]
2026-03-15 - Mobile CSS, Gemini message handling fixes, send_audio logging, audio buffering, and the WebSocket drain task were added.
[DAY 7]
2026-03-15 - Final responsive polish, spacing cleanup, footer/link updates, and the last submission pass closed out the build.
TECH STACK
Google Gemini
Live Multimodal API
Firestore
Session Memory
FastAPI
Python Backend Node
Cloud Run
Serverless Hosting
ADK Toolkit
Agent Framework
WHAT'S NEXT
> tonelens --roadmap
Mobile app (React Native)
Browser extension
Enterprise API
Real-time earpiece mode
Gemini Live Agent Challenge 2026