AI Interview Platform Skill

Overview

智能 AI 面试官平台 - 基于大语言模型的简历分析和模拟面试系统。这是一个全栈应用，集成了简历分析、模拟面试和知识库管理功能，利用 Spring AI 2.0、PostgreSQL + pgvector、Redis 和 React 构建。

技术栈: Java 21, Spring Boot 4.0, Spring AI 2.0, PostgreSQL + pgvector, Redis, React 18.3, TypeScript, Vite, Tailwind CSS

Project Structure

interview-guide/
├── app/                              # 后端应用 (Spring Boot)
│   ├── src/main/java/interview/guide/
│   │   ├── App.java                  # 主启动类
│   │   ├── common/                   # 通用模块
│   │   │   ├── config/               # 配置类
│   │   │   ├── exception/            # 异常处理
│   │   │   └── result/               # 统一响应
│   │   ├── infrastructure/           # 基础设施
│   │   │   ├── export/               # PDF 导出
│   │   │   ├── file/                 # 文件处理
│   │   │   ├── redis/                # Redis 服务
│   │   │   └── storage/              # 对象存储
│   │   └── modules/                  # 业务模块
│   │       ├── interview/            # 面试模块
│   │       ├── jd/                   # 职位描述模块
│   │       ├── knowledgebase/        # 知识库模块
│   │       ├── resume/               # 简历模块
│   │       └── user/                 # 用户模块
│   └── src/main/resources/
│       ├── application.yml           # 应用配置
│       └── prompts/                  # AI 提示词模板
│
├── frontend/                         # 前端应用 (React + TypeScript)
│   ├── src/
│   │   ├── api/                      # API 接口
│   │   ├── components/               # 公共组件
│   │   ├── pages/                    # 页面组件
│   │   ├── types/                    # 类型定义
│   │   └── utils/                    # 工具函数
│   ├── package.json
│   └── vite.config.ts
│
├── docker/                           # Docker 配置文件
├── docker-compose.yml                # Docker Compose 配置
└── README.md

Core Features

1. Resume Management (简历管理)

Multi-format Parsing: Support for PDF, DOCX, DOC, TXT formats
Async Processing: Redis Stream-based async resume analysis with real-time progress tracking
Status Flow: PENDING → PROCESSING → COMPLETED / FAILED
Auto-retry: Built-in failure retry mechanism (max 3 attempts)
Duplicate Detection: Content hash-based duplicate detection
PDF Export: One-click export of AI analysis results to structured PDF reports

2. Mock Interview (模拟面试)

Personalized Questions: AI-generated interview questions based on resume content
Smart Follow-up: Configurable multi-round follow-up questions (default: 1)
Batch Evaluation: Innovative batch evaluation strategy to prevent token overflow
Summary Reports: Secondary aggregation of evaluation results with improvement suggestions
PDF Export: Async generation and export of detailed interview assessment reports

3. Knowledge Base (知识库管理)

Document Processing: Auto-upload, chunking, and async vectorization of PDF, DOCX, Markdown files
RAG Retrieval: Vector database integration for Retrieval-Augmented Generation
Streaming Response: Server-Sent Events (SSE) based typewriter-style streaming responses
Smart Q&A: Knowledge base-driven intelligent Q&A with statistics

4. User Authentication (用户认证)

JWT-based authentication system
Secure password hashing
Session management with Redis

Architecture Patterns

Async Processing Flow

Upload Request → Save File → Send Message to Redis Stream → Return Immediately
                                        ↓
                                Consumer Processes Message
                                        ↓
                            Execute Analysis/Vectorization Task
                                        ↓
                                Update Database Status
                                        ↓
                            Frontend Polls for Latest Status

Technology Stack Decisions

| Component | Choice | Rationale | |-----------|--------|-----------| | Database | PostgreSQL + pgvector | Single database for both relational and vector data, simplified architecture | | Message Queue | Redis Stream | Lightweight alternative to Kafka/RabbitMQ, lower operational overhead | | Object Storage | MinIO (S3-compatible) | Self-hosted, S3 API compatible, easy to switch to cloud providers | | Build Tool | Gradle | Modern build tool with better dependency management | | Frontend Framework | React 18 + TypeScript | Type safety, component reusability, large ecosystem | | Styling | Tailwind CSS 4 | Utility-first CSS framework, rapid UI development |

Development Setup

Prerequisites

| Dependency | Version | Required | |------------|---------|----------| | JDK | 21+ | Yes | | Node.js | 18+ | Yes | | PostgreSQL | 14+ | Yes | | pgvector extension | - | Yes | | Redis | 6+ | Yes | | S3-compatible storage | - | Yes (MinIO recommended) |

Quick Start with Docker (Recommended)

# 1. Clone repository
git clone https://github.com/Snailclimb/interview-guide.git
cd interview-guide

# 2. Configure environment variables
cp .env.example .env
# Edit .env and set your AI_BAILIAN_API_KEY

# 3. Start all services
docker-compose up -d --build

# 4. Access services
# Frontend: http://localhost
# Backend API: http://localhost:8080
# MinIO Console: http://localhost:9001 (minioadmin/minioadmin)

Manual Setup

Backend Setup

# 1. Create database
psql -U postgres
CREATE DATABASE interview_guide;
CREATE EXTENSION vector;

# 2. Set environment variables
export AI_BAILIAN_API_KEY=your_api_key

# 3. Configure application.yml
# Edit app/src/main/resources/application.yml with your database credentials

# 4. Run backend
./gradlew bootRun
# Backend starts at http://localhost:8080

Frontend Setup

cd frontend
pnpm install
pnpm dev
# Frontend starts at http://localhost:5173

Key Configuration

application.yml (Critical Settings):

spring:
  datasource:
    url: jdbc:postgresql://${POSTGRES_HOST:localhost}:${POSTGRES_PORT:5432}/${POSTGRES_DB:interview_guide}
    username: ${POSTGRES_USER:postgres}
    password: ${POSTGRES_PASSWORD:123456}
  
  jpa:
    hibernate:
      ddl-auto: create  # First run: create, then change to update
  
  ai:
    vectorstore:
      pgvector:
        initialize-schema: true  # Dev: true, Prod: false

app:
  interview:
    follow-up-count: ${APP_INTERVIEW_FOLLOW_UP_COUNT:1}
    evaluation:
      batch-size: ${APP_INTERVIEW_EVALUATION_BATCH_SIZE:8}
  storage:
    endpoint: ${APP_STORAGE_ENDPOINT:http://localhost:9000}
    access-key: ${APP_STORAGE_ACCESS_KEY:minioadmin}
    secret-key: ${APP_STORAGE_SECRET_KEY:minioadmin}
    bucket: ${APP_STORAGE_BUCKET:interview-guide}

Common Issues & Solutions

Issue 1: Database Tables Not Created / Data Loss

Problem: JPA ddl-auto configuration issue

Solution:

# First startup - use 'create' to generate tables
jpa:
  hibernate:
    ddl-auto: create

# After tables are created - change to 'update' to preserve data
jpa:
  hibernate:
    ddl-auto: update

Warning: Always change back to update after first run, otherwise all data will be deleted on restart!

Issue 2: Vector Store Table Missing

Problem: When initialize-schema: false, Spring AI won't auto-create vector_store table

Solution:

spring:
  ai:
    vectorstore:
      pgvector:
        initialize-schema: true  # Set to true in dev environment

For production, set to false and manage schema manually.

Issue 3: Resume Analysis Fails

Problem: Analysis stuck or failing

Troubleshooting:

Verify AI_BAILIAN_API_KEY is correctly configured
Check Redis connection and Stream Consumer status
Review backend logs for error messages
Ensure Alibaba Cloud DashScope API key is valid

Issue 4: Resume Stuck in "Processing" State

Problem: Status remains PROCESSING indefinitely

Troubleshooting:

Check Redis connectivity: redis-cli ping
Verify Stream consumer is running
Check backend logs for consumer errors
Restart the application if needed

Issue 5: PDF Export Chinese Character Issues

Problem: Chinese characters not displaying correctly in exported PDFs

Solution:

Font file location: app/src/main/resources/fonts/ZhuqueFangsong-Regular.ttf
Verify iText dependencies are correctly included
Check font loading logs during startup

API Endpoints Overview

Resume Module

POST /api/resumes/upload - Upload resume for analysis
GET /api/resumes/{id} - Get resume analysis details
GET /api/resumes/{id}/report - Export PDF report
DELETE /api/resumes/{id} - Delete resume

Interview Module

POST /api/interviews/start - Start mock interview session
POST /api/interviews/{id}/answer - Submit answer
GET /api/interviews/{id} - Get interview details
POST /api/interviews/{id}/export - Export interview report

Knowledge Base Module

POST /api/knowledge/documents - Upload document
POST /api/knowledge/chat - Chat with knowledge base (SSE)
GET /api/knowledge/statistics - Get knowledge base statistics

User Module

POST /api/auth/register - User registration
POST /api/auth/login - User login
GET /api/users/profile - Get user profile

Deployment Guide

Docker Deployment

The project provides complete Docker support with one-command deployment:

# Start all services (frontend, backend, database, middleware)
docker-compose up -d --build

# View service status
docker-compose ps

# View backend logs
docker-compose logs -f app

# Stop all services
docker-compose down

# Clean up unused images
docker image prune -f

Service Ports

| Service | URL | Default Credentials | Description | |---------|-----|-------------------|-------------| | Frontend | http://localhost | - | User interface | | Backend API | http://localhost:8080 | - | REST API / Swagger | | MinIO Console | http://localhost:9001 | minioadmin/minioadmin | Object storage management | | MinIO API | localhost:9000 | - | S3-compatible API | | PostgreSQL | localhost:5432 | postgres/password | Database with pgvector | | Redis | localhost:6379 | - | Cache & message queue |

Environment Variables (.env)

Required variables:

AI_BAILIAN_API_KEY=your_api_key_here
AI_MODEL=qwen-plus  # Optional, default: qwen-plus
APP_INTERVIEW_FOLLOW_UP_COUNT=1  # Optional, default: 1
APP_INTERVIEW_EVALUATION_BATCH_SIZE=8  # Optional, default: 8

Testing

Backend Testing

./gradlew test

Frontend Testing

cd frontend
pnpm build  # Build for production

Contributing Guidelines

Fork the repository
Create a feature branch (git checkout -b feature/amazing-feature)
Commit your changes (git commit -m 'Add amazing feature')
Push to the branch (git push origin feature/amazing-feature)
Open a Pull Request

License

AGPL-3.0 License - If you provide the software as a network service, you must make the modified source code available to users.

Use Cases

| User Role | Use Case | |-----------|----------| | Job Seekers | Upload resumes for AI analysis, practice mock interviews | | HR/Recruiters | Batch analyze resumes, evaluate candidate capabilities | | Training Organizations | Provide interview training services, manage knowledge bases |

Additional Resources

Architecture Diagram: See README.md for detailed system architecture
Tutorial: Comprehensive guide available at JavaGuide Interview Guide (paid content)
Issue Tracker: Report bugs and request features via GitHub Issues

Skill Usage Instructions

This skill can be used to:

Understand the project structure and architecture
Set up development environment (Docker or manual)
Troubleshoot common issues
Deploy the application using Docker
Configure AI models and external services
Implement new features following existing patterns
Debug async processing flows
Extend knowledge base functionality

When working with this codebase:

Always check application.yml for configuration issues first
Verify Redis and PostgreSQL connectivity for async task problems
Use Docker Compose for quick local development setup
Follow the existing module structure when adding new features
Test PDF exports with Chinese characters carefully
Monitor Redis Stream consumers for stuck tasks

InterLyt

AI Interview Platform Skill

Overview

Project Structure

Core Features

1. Resume Management (简历管理)

2. Mock Interview (模拟面试)

3. Knowledge Base (知识库管理)

4. User Authentication (用户认证)

Architecture Patterns

Async Processing Flow

Technology Stack Decisions

Development Setup

Prerequisites

Quick Start with Docker (Recommended)

Manual Setup

Backend Setup

Frontend Setup

Key Configuration

Common Issues & Solutions

Issue 1: Database Tables Not Created / Data Loss

Issue 2: Vector Store Table Missing

Issue 3: Resume Analysis Fails

Issue 4: Resume Stuck in "Processing" State

Issue 5: PDF Export Chinese Character Issues

API Endpoints Overview

Resume Module

Interview Module

Knowledge Base Module

User Module

Deployment Guide

Docker Deployment

Service Ports

Environment Variables (.env)

Testing

Backend Testing

Frontend Testing

Contributing Guidelines

License

Use Cases

Additional Resources

Skill Usage Instructions