---

🚀 BitQuant by OpenGradient

BitQuant is a open-source AI agent framework for building quantitative AI agents. It leverages specialized models for ML-powered analytics, trading, portfolio management, and more—all through a natural language interface. BitQuant exposes a REST API that turns user inputs like "What is the current risk profile on Bitcoin?" or "Optimize my portfolio for maximum risk-adjusted returns" into actionable insights.

---

📑 Table of Contents

• Features
• Architecture
• Installation
• Usage
• Configuration
• Integrations
• Testing
• Deployment
• Contributing
• License
• Contact

---

✨ Features

• 🤖 Build and deploy quantitative AI agents for analytics, trading, and portfolio management
• 🧠 Natural language interface for complex financial queries
• 🔌 Modular architecture with agent and tool plug-ins
• 📈 Real-time crypto analytics and risk profiling
• 🌐 REST API for seamless integration
• ⚡ Fast setup and extensible codebase

🏗️ Architecture

agent/      # Agent logic and tool definitions
api/        # Server API input/output types
onchain/    # Classes for on-chain data (tokens, pools, etc.)
server/     # Flask server exposing the API
static/     # Static assets for web interface
subnet/     # Bittensor Subnet-related functionality
templates/  # LLM prompt templates for agent
testclient/ # Client for testing the API
testutils/  # Utility functions for testing

Agents

• Analytics Agent: Handles crypto analytics (price trends, risks, trending tokens, etc.)
• Investment Agent: Helps users select lending/AMM pools to maximize returns on Solana

The router in server.py decides which agent to use for each user query.

⚙️ Installation

make venv
source venv/bin/activate
make install

🚦 Usage

Environment Variables

Before running the service, you need to set up the following environment variables. Create a .env file in the root directory with the following variables:

Required Environment Variables

Firebase Authentication:

FIREBASE_PROJECT_ID=your_firebase_project_id
FIREBASE_PRIVATE_KEY_ID=your_private_key_id
FIREBASE_PRIVATE_KEY="-----BEGIN PRIVATE KEY-----\nYour private key here\n-----END PRIVATE KEY-----"
FIREBASE_CLIENT_EMAIL=your_service_account_email
FIREBASE_CLIENT_ID=your_client_id
FIREBASE_CLIENT_X509_CERT_URL=your_cert_url

Solana RPC:

SOLANA_RPC_URL=your_solana_rpc_endpoint

API Keys:

OPENROUTER_API_KEY=your_openrouter_api_key
GEMINI_API_KEY=your_gemini_api_key
COINGECKO_API_KEY=your_coingecko_api_key

AWS (for DynamoDB):

AWS_ACCESS_KEY_ID=your_aws_access_key
AWS_SECRET_ACCESS_KEY=your_aws_secret_key
AWS_REGION=your_aws_region

Optional Environment Variables

Datadog Monitoring:

DD_API_KEY=your_datadog_api_key
DD_APP_KEY=your_datadog_app_key
DD_HOSTNAME=your_hostname

Cloudflare Turnstile (for CAPTCHA):

CLOUDFLARE_TURNSTILE_SECRET_KEY=your_turnstile_secret_key

Environment:

ENVIRONMENT=development

Running the Service

1. Create your .env file with the required variables:

   # Copy and edit the environment variables above

2. Build the server:

   make docker

3. Start the server:

   make prod

You can also try BitQuant instantly on the production server.

---

💡 Sample Questions

Here are some example queries you can try with BitQuant:

🏦 DeFi Interactions

• Which protocols are delivering the best risk-adjusted yields right now?
• What's my potential impermanent loss risk if I provide liquidity to the USDC-SOL pool under different market scenarios?
• Calculate a comprehensive risk score for the top 5 Solana DeFi protocols based on TVL trends, code audits, and historical performance
• Compare the TVL growth, volatility, and stability metrics for Kamino vs Orca vs Raydium
• Which lending protocols have maintained the most stable yields over the past 3 months?

📊 Portfolio Analytics

• Can you analyze my portfolio's rolling volatility and identify which assets are contributing most to risk?
• How do the volatility trends of my top portfolio assets compare over the last 90 days?
• Show me the correlation between my holdings and provide insights on how to better diversify?
• What's my current portfolio risk assessment and how can I optimize for a better risk-return ratio?
• What's the maximum drawdown for my current portfolio and how does it compare to market benchmarks?

📈 Market Insights

• Based on current volatility trends and price patterns, what phase of the market cycle are we likely in?
• Based on historical data, what's the volatility forecast for BTC and ETH in the coming month?

---

🧑‍💻 Types of Agents You Can Build

BitQuant is designed to support a wide range of quantitative and DeFi-focused AI agents. Out of the box, the framework includes:

1. Analytics Agent

• Purpose: Provides deep analytics on portfolios, tokens, protocols, and market trends.
• Capabilities:
  • Analyze portfolio volatility, drawdowns, and diversification
  • Evaluate token and protocol risks
  • Track TVL, yield, and performance metrics
  • Identify trends and generate actionable market insights
• Example Use Cases:
  • "Analyze my portfolio’s risk profile."
  • "Show TVL trends for Solana DeFi protocols."

2. Investor Agent

• Purpose: Helps users find and act on yield opportunities and optimize DeFi strategies.
• Capabilities:
  • Recommend optimal pools and lending opportunities
  • Compare APRs, TVL, and risk across protocols
  • Guide users through liquidity provision, lending, and yield farming
• Example Use Cases:
  • "Which pools offer the best stablecoin yields?"
  • "Compare Kamino and Orca for USDC/SOL."

3. Custom Agents

• Purpose: The framework is extensible—developers can build agents for:
  • Automated trading strategies
  • On-chain data monitoring and alerting
  • NFT analytics
  • Cross-chain portfolio management
  • Any custom DeFi or analytics workflow

Tip: Agents are modular and can be combined, routed, or extended to suit your specific use case. See the agent/ directory and templates for examples and customization.

---

🛠️ Configuration

• All configuration is handled via the .env file, which you can generate from .env.example.
• Fill in all required secrets and keys as described in .env.example.

🔌 Integrations

• REST API: Exposes endpoints for agent interaction
• Bittensor Subnet: For decentralized compute
• Custom LLM Prompts: In templates/

🧪 Testing

To run all tests:

make test

🚀 Deployment

Build and run in production:

make docker
make prod

🤝 Contributing

Contributions are welcome! Please open issues or pull requests for features, bugs, or documentation improvements.

1. Fork the repo
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.

💬 Contact

• Powered by OpenGradient
• Try Bitquant at bitquant.io
• OpenGradient Documentation: OpenGradient Docs
• Join the discussion: GitHub Discussions
• Support: BitQuant Discord

---

Empowering next-gen quantitative AI agents with OpenGradient.