Qiumi - Indonesia-Australia Economic Explorer

Qiumi provides AI-powered insights into the Indonesia–Australia economic landscape, covering policy, investment, and strategic decision-making. It is built using the Swift programming language (Hummingbird) and Svelte, powered by Llama 4 (Scout) on Vertex AI.

Features

• AI-powered insights using Llama 4 (Scout) on Vertex AI
• Real-time analysis of Indonesia-Australia economic partnerships
• Industry sector collaboration insights
• Investment and policy analysis
• RESTful API endpoints
• CORS-enabled for web frontend integration

Getting Started

Prerequisites

• Swift 6.2 or later
• macOS 14 or later
• Google Cloud account with Llama 4 API access
• gcloud CLI installed and authenticated

Setup

1. Clone the repository

   git clone https://github.com/harmoniousmoss/qiumi.git
   cd qiumi

2. Configure environment variables

   Create a .env file in the project root:

   touch .env

   Add your Google Cloud configuration to .env:

   PROJECT_ID=
   REGION=
   ENDPOINT=
   LLAMA_MODEL=

   Refer to Google Cloud Vertex AI Llama documentation for the appropriate values for your setup.

   Note: The .env file is gitignored and will not be committed to the repository.

3. Authenticate with Google Cloud

   Make sure you have gcloud CLI installed and authenticated:

   gcloud auth login
   gcloud config set project YOUR_PROJECT_ID

   The application will automatically use your gcloud credentials to access the Llama 4 API.

Running the Application

Build and run the application:

swift build
swift run

The server will start on http://127.0.0.1:8080

You should see:

2025-11-06T16:52:02+0700 info Hummingbird: [HummingbirdCore] Server started and listening on 127.0.0.1:8080

API Usage

Get Economic Insights

Endpoint: POST /insights

Request Body:

{
  "question": "How are Indonesia and Australia partnering in the mining industry sector?"
}

Response:

{
  "answer": "Indonesia and Australia have established significant partnerships in the mining sector..."
}

Example using curl:

curl -X POST http://127.0.0.1:8080/insights -H "Content-Type: application/json" -d '{"question": "What are the key investment opportunities between Indonesia and Australia?"}'

Important: Make sure the entire command is on one line, or if using line breaks with \, ensure there are no line breaks inside the JSON string value.

Example Questions

• "How are Indonesia and Australia partnering in the mining industry sector?"
• "What are the main trade agreements between Indonesia and Australia?"
• "What investment opportunities exist in renewable energy between both countries?"
• "How do Indonesia and Australia collaborate on agricultural exports?"
• "What are the key strategic decisions affecting Indonesia-Australia economic relations?"

Architecture

• Configuration: Environment variable loader for secure credential management
• LlamaService: HTTP client wrapper for Google Cloud Vertex AI Llama 4 API
• API Endpoints: RESTful endpoints with comprehensive error handling
• Logging: Detailed request/response logging for debugging

Technologies

• Swift - Programming language
• Hummingbird - Web framework
• Llama 4 (Scout) - AI model via Google Cloud Vertex AI
• AsyncHTTPClient - HTTP client
• Svelte - Frontend framework (separate repository)

Security

• The .env file containing sensitive credentials is excluded from version control
• All API requests use OAuth 2.0 tokens via Google Cloud CLI
• Environment variables are loaded at runtime from .env file

Contributing

When contributing, ensure:

1. Never commit the .env file
2. Update .env.example with any new configuration variables (without values)
3. Test API endpoints thoroughly before submitting changes

License

This project is built using the Hummingbird framework.