A full-stack application for managing and analyzing customer service tickets for Mihoyo. The system provides automated ticket clustering, issue summarization, and FAQ generation capabilities.
- Ticket Clustering: Automatically groups similar customer service tickets using advanced clustering algorithms
- Issue Summarization: Generates concise summaries of ticket clusters to identify common patterns
- FAQ Generation: Automatically creates FAQ entries based on common issues
- Interactive UI: React-based frontend for easy ticket management and analysis
- Real-time Processing: Handles ticket processing through Google Cloud PubSub
- Data Analytics: Leverages BigQuery for efficient data analysis
.
├── mihoyo-cs-tickets-ui/ # React TypeScript frontend
├── sql/ # SQL scripts for data processing
├── bq_handler.py # BigQuery integration handler
├── cluster_issue.py # Ticket clustering logic
├── config.toml # Configuration file
├── main.py # Main application entry point
├── pubsub_handler.py # PubSub message handling
└── summary_issue.py # Issue summarization logic
- Python 3.x
- Node.js and npm
- Google Cloud Platform account with BigQuery and PubSub enabled
-
Backend Setup
# Install Python dependencies pip install -r requirements.txt # Configure the application # Edit config.toml with your settings
-
Frontend Setup
cd mihoyo-cs-tickets-ui npm install
-
Start the Backend
uvicorn main:app --reload --port 8000
-
Start the Frontend
cd mihoyo-cs-tickets-ui npm start
The application will be available at http://localhost:3000
The backend is built with Python and consists of several key components:
main.py: Application entry point and API endpointsbq_handler.py: Handles BigQuery operationscluster_issue.py: Implements ticket clustering logicpubsub_handler.py: Manages PubSub message processingsummary_issue.py: Handles ticket summarization
The frontend is built with React and TypeScript, located in the mihoyo-cs-tickets-ui directory:
/src/components: React components/src/services: API service integrations/src/types: TypeScript type definitions
SQL scripts in the sql/ directory handle:
- View creation
- Issue summarization
- Embedding generation
- FAQ generation
Edit config.toml to configure:
- Database connections
- GCP settings
- Application parameters
- API endpoints
[Add your license information here]



