Question Marking

A batch AI grading engine designed for throughput, stability, and large-scale processing. The project focuses on concurrent grading, caching, retries, validation, and performance monitoring rather than one-off demo calls.

What this project does

• Processes folders of question-answer JSON files in batch
• Calls AI services to generate grading results
• Caches reusable intermediate outputs
• Retries failed requests automatically
• Tracks performance metrics during long-running jobs
• Produces structured result summaries for downstream analysis

Why it matters

The value of this repository is engineering depth. It shows how to turn grading logic into a repeatable batch system with:

• concurrency control
• caching
• retries
• validation
• monitoring

Main files

• marking.py: main grading workflow
• performance_monitor.py: runtime metrics and monitoring
• process_output.py: output processing and result formatting
• improved_path_extraction.py: path and file handling helpers
• config.json: runtime configuration

Repository structure

question-marking/
├── marking.py
├── performance_monitor.py
├── process_output.py
├── improved_path_extraction.py
├── config.json
├── requirements.txt
└── test_*.py

Quick start

pip install -r requirements.txt
cp .env.example .env
python marking.py

Key capabilities

• Concurrent batch grading
• Cache and retry mechanisms
• Configurable processing pipeline
• Performance monitoring
• Validation and fault tolerance

Notes

This public repository is a cleaned release version. Local caches, logs, evaluation outputs, runtime folders, and private environment files were excluded.