Skip to content

About

Voice + vision AI agent that lets artists prove their work is human-made. Built for the Google Gemini Live Agent Challenge ($25K). Powered by Gemini Live API + Solana.

Resources

Stars

2 stars

Watchers

0 watching

Forks

Latest commit

 

History

11 Commits

Folders and files

Repository files navigation

AetherProof

Voice + vision AI agent that lets artists prove their work is human-made.

Built for the Google Gemini Live Agent Challenge — $25,000 Grand Prize | Deadline: March 16, 2026

Day 1 of 19 Built with Gemini Solana


What is AetherProof?

Artists are losing income and recognition to AI-generated look-alikes. AetherProof lets a human artist speak to an agent, show their canvas, and receive a verifiable on-chain certificate proving the work is human-made — in real time.

How it works:

  1. Artist opens AetherProof and starts a live session
  2. They speak naturally: "I'm painting this landscape. Watch me work."
  3. Gemini Live API analyzes the audio + live canvas feed
  4. Agent confirms human creation patterns (brushstroke analysis, real-time narration, process verification)
  5. A Human Touch Certificate is minted on Solana devnet — timestamped, immutable, shareable

Tech Stack

Layer Technology
Frontend React + TypeScript
Real-time audio/vision Gemini Live API via Google GenAI SDK
Agent framework Google ADK (Agent Development Kit)
Backend Node.js + TypeScript on Google Cloud Run
Database Firestore (session + certificate store)
On-chain proof Solana (@solana/web3.js, devnet)
Deployment Google Cloud Run + Cloud Build

Project Structure

aetherproof/
├── backend/          # Node.js/TypeScript — Gemini Live API + Cloud Run
│   ├── src/
│   │   ├── index.ts  # Express server entry point
│   │   ├── gemini.ts # Gemini Live API stream handler
│   │   └── solana.ts # Certificate minting on Solana
│   └── package.json
├── frontend/         # React/TypeScript — live audio UI
│   ├── src/
│   │   ├── App.tsx
│   │   ├── components/
│   │   │   ├── AudioWaveform.tsx
│   │   │   └── CertificateCard.tsx
│   │   └── hooks/
│   │       └── useGeminiStream.ts
│   └── package.json
├── docs/
│   └── architecture.md  # Architecture diagram (Devpost requirement)
└── README.md

Getting Started

# Clone
git clone https://github.com/Mellowambience/aetherproof.git
cd aetherproof

# Backend
cd backend
npm install
cp .env.example .env  # Add your GEMINI_API_KEY
npm run dev

# Frontend (new terminal)
cd frontend
npm install
npm start

Environment variables:

GEMINI_API_KEY=your_gemini_api_key_here
SOLANA_NETWORK=devnet
PORT=8080

Submission Checklist

  • Text description on Devpost
  • Public GitHub repo with spin-up instructions (this README)
  • GCP Cloud Run deployment proof
  • Architecture diagram (docs/architecture.md)
  • Demo video <4 minutes
  • Blog post with #GeminiLiveAgentChallenge
  • Google Developer Group membership

Build Log

Day Date Milestone
1 Feb 25 Repo created, monorepo scaffold, Gemini Live API stub
... ... ...

Built by @1Aether1Rose1 — mellowambience.github.io

About

Voice + vision AI agent that lets artists prove their work is human-made. Built for the Google Gemini Live Agent Challenge ($25K). Powered by Gemini Live API + Solana.

Resources

Stars

2 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages