AI-powered web application that analyzes a user's CV against job descriptions, calculates semantic match scores, identifies missing skills, and generates personalized improvement suggestions.
Beyond its functional goal, this project is designed as a production-oriented backend system, with a strong focus on reliability, failure handling, and supportability.
- CV upload (PDF / DOC / DOCX)
- Automatic text extraction
- AI-based semantic matching between CV & Job Description
- Match score calculation (0–100%)
- Missing skill detection
- Personalized improvement suggestions
- JWT-based authentication & authorization
- User-specific CV & analysis history
- Dashboard-style UI
This project was built with real-world application support and production operations in mind:
- Integration with an external AI microservice and handling service failures
- Clear separation between API layer and AI processing layer
- Structured logging for troubleshooting and incident investigation
- Secure authentication and protected endpoints
- Database migrations managed via Flyway
- Environment-based configuration for different runtime setups
- Designed to remain operational and debuggable during partial failures
These aspects simulate realistic production scenarios where external services, databases, or configurations may fail and require structured investigation.
- Java 17
- Spring Boot
- Spring Security + JWT
- JPA / Hibernate
- PostgreSQL
- Flyway
- Docker-ready architecture
- React
- TypeScript
- Axios
- Modern dark UI
- Python-based AI microservice
- Sentence embeddings
- Semantic similarity matching
- Skill extraction and suggestion engine
Client (React)
→ Spring Boot REST API
→ AI Microservice (Python)
→ PostgreSQL
This architecture enables isolation of failures and easier operational support.
- User registers and logs in
- JWT token issued on authentication
- Token attached to API requests for protected endpoints
mvn spring-boot:run