AI Text Processing System

A comprehensive LLM text processing system integrating text polishing, AI detection, and plagiarism reduction features, developed during my internship at Beijing ZaiwenAI Technology Co., Ltd.

Project Overview

This project represents a complete full-stack development cycle, from product design and tech stack selection to cloud deployment. The system provides intelligent text processing capabilities using state-of-the-art language models and detection algorithms.

Technical Stack

• Backend: FastAPI with Celery+Redis asynchronous task queue
• Frontend: HTML5+JavaScript native version and Vue.js component-based refactored version
• AI/ML: DeepSeek-R1 API, GPTZero detection API, custom-trained RoBERTa/BERT models
• Deployment: Render cloud platform with GitHub Pages static hosting
• CI/CD: Automated deployment pipeline

Key Features

• Text Polishing: Advanced text improvement using LLM APIs
• AI Detection: Integration with GPTZero and custom-trained models for AI-generated content detection
• Plagiarism Reduction: Intelligent text rewriting to reduce similarity scores
• Performance Optimization: 65%+ improvement in processing efficiency through asynchronous task queues
• Responsive Design: Cross-platform compatibility with modern web standards

Architecture Highlights

• RESTful API backend designed for scalability
• Asynchronous task processing for handling concurrent requests
• Modular frontend architecture allowing for easy maintenance and updates
• Cloud-native deployment with automated CI/CD pipeline

Performance Metrics

• Efficiency Improvement: 65%+ increase in processing speed
• Scalability: Handles multiple concurrent requests efficiently
• Reliability: Robust error handling and task queue management

Links

• Source Code: GitHub Repository
• Live Demo: Demo Site
• Documentation: Comprehensive README with setup instructions

Impact

This project demonstrates proficiency in full-stack development, cloud deployment, and AI/ML integration. It showcases the ability to manage complete product lifecycles and implement modern web technologies effectively.