EchoMind is a decentralized AI assistant built on the OpenGradient network. It leverages Verifiable Inference through Trusted Execution Environments (TEEs) to provide secure, private, and transparent AI interactions. By integrating MemSync, EchoMind maintains a persistent, decentralized memory of user technical interests and past conversations.
- Verifiable AI Inference: Powered by the OpenGradient SDK using the
GEMINI_2_5_FLASHmodel running in a TEE. - Decentralized Long-term Memory: Uses MemSync API to store and retrieve user-specific context, ensuring the AI "remembers" you without centralized data silos.
- Secure Backend: Built with FastAPI and orchestrated on a VPS with screen-managed sessions for high availability.
- Modern Frontend: A responsive UI built with Lovable for seamless user interaction.
| Component | Technology |
|---|---|
| Language | Python 3.12+ |
| API Framework | FastAPI |
| Blockchain/AI | OpenGradient SDK (Base Sepolia) |
| Memory Layer | MemSync (Decentralized Vector Storage) |
| Server | Uvicorn |
| Orchestration | Linux Screen |
- Clone the repository:
git clone [https://github.com/mrnetwork0001/echomind.git](https://github.com/mrnetwork0001/echomind.git) cd echomind - Set up a Virtual Environment:
python3 -m venv venv
source venv/bin/activate- Install Dependencies:
pip install fastapi uvicorn opengradient requests certifi- Environment Variables:
Create an
.envfile or export your keys:
export OG_PRIVATE_KEY="your_private_key"
export MEMSYNC_API_KEY="your_memsync_key"The backend runs on Port 8000.
Endpoint: POST /chat
Body:
{
"user_id": "unique_user_id",
"message": "Hello EchoMind!"
}This project utilizes the Permit2 protocol via OpenGradient to handle gasless AI inferences securely. Every response is processed on-chain/TEE, providing a cryptographic guarantee of the model used.
MIT License