This directory contains Bruno API collections for testing VRecommendSystem endpoints.

Bruno is a fast, git-friendly, open-source API client. This collection provides ready-to-use API requests for all VRecommendSystem endpoints.

Download and install Bruno from: https://www.usebruno.com/

1. Launch Bruno
2. Click "Open Collection"
3. Navigate to this directory (vrecom_api/)
4. Click "Select Folder"

The collection uses the following default URLs:

• API Server: http://localhost:2030
• AI Server: http://localhost:9999

Update the URLs in individual requests if your ports differ.

vrecom_api/
├── api_server/              # API Server endpoints
│   ├── auth/               # Authentication (Google OAuth)
│   │   ├── login_google.bru
│   │   ├── check_auth_status.bru
│   │   └── logout.bru
│   ├── recommendations/    # Get recommendations
│   │   └── recommend.bru
│   └── ping.bru           # Health check
│
└── ai_server/              # AI Server endpoints
    ├── models/             # Model management
    │   ├── create_model.bru
    │   ├── list_models.bru
    │   ├── get_model_info.bru
    │   └── delete_model.bru
    ├── tasks/              # Task scheduler
    │   ├── add_model_task.bru
    │   ├── list_tasks.bru
    │   ├── remove_model_task.bru
    │   ├── rename_task.bru
    │   └── update_task_interval.bru
    ├── data_chefs/         # Data pipelines
    │   ├── create_data_chef_from_csv.bru
    │   ├── create_data_chef_from_sql.bru
    │   ├── list_data_chefs.bru
    │   ├── get_data_chef.bru
    │   └── delete_data_chef.bru
    ├── scheduler/          # Scheduler control
    │   ├── stop_scheduler.bru
    │   └── restart_scheduler.bru
    ├── metrics/            # System metrics
    │   ├── get_total_running_tasks.bru
    │   ├── get_total_activated_tasks.bru
    │   ├── get_total_activating_models.bru
    │   ├── get_scheduler_status.bru
    │   └── get_server_logs.bru
    ├── recommendations/    # Get recommendations
    │   └── recommend.bru
    ├── health.bru         # Health check
    └── main.bru           # Root endpoint

# Start all services with Docker
./docker-start.sh up

# Or start individually
cd backend/api_server && go run main.go
cd backend/ai_server && poetry run server

• API Server: GET http://localhost:2030/api/v1/ping
• AI Server: GET http://localhost:9999/api/v1/health

Use ai_server/models/create_model.bru:

{
  "model_id": "my_model",
  "model_type": "svd",
  "model_config": {
    "n_factors": 100,
    "n_epochs": 20,
    "lr_all": 0.005,
    "reg_all": 0.02
  }
}

Use ai_server/data_chefs/create_data_chef_from_csv.bru or SQL variant.

Use ai_server/tasks/add_model_task.bru:

{
  "task_name": "daily_training",
  "model_id": "my_model",
  "interactions_data_chef_id": "interactions_chef",
  "interval": "daily"
}

Use ai_server/recommendations/recommend.bru:

GET /api/v1/recommend/user123/my_model/10

The API Server uses Google OAuth for authentication:

1. Access auth/login_google.bru to initiate OAuth flow
2. Complete Google authentication in browser
3. Session cookie will be stored automatically
4. Use auth/check_auth_status.bru to verify session

The AI Server requires authentication token in request headers:

Authorization: Bearer <your-token>

POST /api/v1/create_model
{
  "model_id": "svd_model",
  "model_type": "svd",
  "model_config": {
    "n_factors": 100,
    "n_epochs": 20
  }
}

POST /api/v1/create_model
{
  "model_id": "nmf_model",
  "model_type": "nmf",
  "model_config": {
    "n_factors": 50,
    "n_epochs": 15
  }
}

POST /api/v1/create_data_chef_from_csv
{
  "data_chef_id": "csv_interactions",
  "file_path": "/data/interactions.csv",
  "user_column": "user_id",
  "item_column": "item_id",
  "rating_column": "rating"
}

POST /api/v1/add_model_task
{
  "task_name": "hourly_training",
  "model_id": "my_model",
  "interactions_data_chef_id": "interactions_chef",
  "interval": "hourly"
}

Available intervals for scheduled tasks:

• hourly: Every hour
• daily: Every day at midnight
• weekly: Every week
• monthly: Every month
• Custom cron expressions

• Total running tasks: GET /api/v1/get_total_running_tasks
• Active models: GET /api/v1/get_total_activating_models
• Scheduler status: GET /api/v1/get_scheduler_status
• Server logs: GET /api/v1/get_server_logs?limit=100

Access Prometheus metrics at: http://localhost:9090

• Verify services are running: ./docker-start.sh status
• Check port configuration in .env
• View service logs: ./docker-start.sh logs <service_name>

• Ensure you've completed OAuth flow for API Server
• Check token is present in request headers for AI Server
• Verify Redis is running for session storage

• Check data chef is properly configured
• Verify CSV/database connection
• Review AI Server logs: ./docker-start.sh logs ai_server

• Bruno Documentation
• Project README
• Docker Setup Guide
• System Architecture

1. Environment Variables: Use Bruno's environment feature to switch between dev/staging/prod
2. Collections: Organize related requests into folders
3. Pre-request Scripts: Add authentication token generation if needed
4. Tests: Add response assertions for automated testing
5. Version Control: Bruno collections are git-friendly - commit them!