CryptoAgent: Cryptocurrency Analysis Platform

CryptoAgent is an advanced cryptocurrency analysis platform that provides real-time price data, generates visual charts, and offers financial investment advice for various cryptocurrencies. The platform utilizes AI agents powered by Google's Vertex AI to deliver comprehensive crypto insights.

Features

• Finance Advisor Agent: Get personalized investment advice for cryptocurrencies including risk assessment, potential returns, and platform recommendations
• Graph Generator Agent: Generate visual charts showing cryptocurrency price trends and performance metrics
• Crypto Price Agent: Fetch real-time price data including current price, market cap, and price changes over different time periods

Live Application

The application is deployed and available at:

https://cryptoirwa-frontend-1088335055755.asia-southeast1.run.app/

API Documentation

Access the interactive API documentation:

http://127.0.0.1:8000/docs

Setup Instructions

1. Clone the Repository

git clone https://github.com/saiful247/CryptoIRWA.git
cd CryptoAgent

2. Create Virtual Environment

python -m venv venv

3. Activate Virtual Environment

Windows:

venv\Scripts\activate

macOS/Linux:

source venv/bin/activate

4. Install Dependencies

pip install -r requirements.txt

5. Environment Configuration

Create a .env file in the project root with the following configuration:

# Vertex AI Configuration
GOOGLE_APPLICATION_CREDENTIALS=path/to/your/service-account-key.json
PROJECT_ID=your-gcp-project-id
LOCATION=your-gcp-region

# API Keys
COINMARKETCAP_API_KEY=your_coinmarketcap_api_key

# JWT Configuration
JWT_SECRET_KEY=your_jwt_secret_key
JWT_ALGORITHM=HS256
JWT_ACCESS_TOKEN_EXPIRE_MINUTES=30

# GCP Storage
GCS_BUCKET_NAME=your-gcs-bucket-name

# Firestore Configuration
FIRESTORE_COLLECTION_NAME=your-firestore-collection

6. Run the Application

uvicorn app.main:app --reload

The application will be available at http://127.0.0.1:8000.

Technologies Used

• FastAPI - Web framework
• Google Vertex AI - AI agent framework and models
  • Imagen 4.0 (imagen-4.0-generate-001) - AI image generation for NFT creation
  • Gemini 2.5 Flash (gemini-2.5-flash) - Advanced language model for analysis and advice
• LangGraph - Agent workflow orchestration and state management
• Google Cloud Run - Serverless deployment platform
• Google Cloud Storage (GCS) - File storage
• Google Firestore - NoSQL document database
• Docker - Containerization
• JWT - Authentication and authorization
• CoinMarketCap API - Cryptocurrency data
• YFinance - Financial data

Architecture

• Vertex AI Agentic System: Uses Google's Vertex AI for intelligent cryptocurrency analysis
• GCP Cloud Run: Serverless deployment for scalable application hosting
• Docker: Containerized deployment for consistency across environments
• GCP Bucket: Cloud storage for static assets and generated charts
• Firestore: NoSQL database for storing user data and analysis results
• JWT Authentication: Secure token-based authentication system

Resources

---

Docker & Deployment

Docker Commands

# Build the Docker image
docker build -t cryptoagent-app .

# Run the container locally
docker run -d --name cryptoagent-container -p 8000:8000 --env-file .env cryptoagent-app

# Stop and remove the container
docker stop cryptoagent-container
docker rm cryptoagent-container

# View container logs
docker logs -f cryptoagent-container

Deployment (GCP Cloud Run)

The application is deployed on Google Cloud Run with the following features:

• Automatic scaling based on traffic
• Integrated with GCP services (Vertex AI, Firestore, Cloud Storage)
• Secure JWT-based authentication
• Docker containerization for consistent deployment

Environment Variables for Deployment

Ensure the following environment variables are set in Cloud Run:

• PROJECT_ID
• LOCATION
• COINMARKETCAP_API_KEY
• JWT_SECRET_KEY
• JWT_ALGORITHM
• GCS_BUCKET_NAME
• FIRESTORE_COLLECTION_NAME

Service Account Configuration

• Do NOT upload service account credentials to Cloud Run
• Instead, select the service account that your Vertex AI uses in the Cloud Run service configuration
• GCP Cloud Run will automatically detect and use the selected service account
• Required permissions for the service account: Vertex AI User, Firestore User, Storage Object Admin

Other Issues & Tips

• In .env, do not wrap values (like GOOGLE_APPLICATION_CREDENTIALS, ALGORITHM, API keys) in double quotes (""). This causes Docker build issues.
• Keep your Dockerfile as simple as possible.
• If deployment fails, check Cloud Run logs for details.

Migration Notes

• Migrated from Autogen to Vertex AI: Vertex AI provides better integration with GCP services, more model variety, and easier deployment for production environments
• Cloud-Native Architecture: Fully integrated with Google Cloud Platform for scalability and reliability
• Secure Authentication: Implemented JWT-based authentication for secure API access