Role-Based Authentication
Secure authentication flow with different permissions and protected experiences for applicants, reviewers and managers.
FeaturedCompletedFull StackAI
AI Smart ApplicationSystem
An AI-powered recruitment platform designed to simplify candidate applications, CV analysis and decision-making through secure, role-based digital workflows.
3
User roles
Full Stack
Application type
AI
CV analysis
Deployed
Production workflow
Project overview
More than a recruitment dashboard
The platform connects applicants, reviewers and managers through one centralized system.
AI Smart Application System is a full-stack recruitment platform created to improve how applications are submitted, reviewed and evaluated.
Applicants can create an account, upload their CV and apply through a structured digital process. Reviewers and managers can access candidate information, inspect applications and use AI-generated analysis to support their decisions.
The project was developed as my master's thesis and combines frontend engineering, backend development, database design, authentication, AI integration and cloud deployment.
My contribution
End-to-end development
Product and interface planning
Frontend implementation
Backend REST API development
MongoDB database integration
Authentication and authorization
AI service integration
Frontend and backend deployment
Problem and solution
Turning a slow manual process into a digital workflow
The product was designed around a practical recruitment problem rather than a collection of disconnected features.
01
The problem
Candidate evaluation can become fragmented and time-consuming
- Applications may arrive through different channels
- Manual CV evaluation requires significant time
- Candidate information can be difficult to organize
- Applicants may have limited visibility into their status
- Different recruitment roles require different access levels
02
The solution
One platform for applications, analysis and decisions
Centralized candidate and application management
AI-supported CV analysis and structured insights
Secure role-based dashboards
Clear application status tracking
Protected backend routes and authenticated workflows
Core functionality
Features built around real user roles
Each part of the system supports a specific step of the recruitment workflow.
AI-Powered CV Analysis
Uploaded CVs are analyzed by AI to generate structured insights and support faster candidate evaluation.
Application Management
Managers and reviewers can access applications, inspect candidate information and manage recruitment decisions.
CV File Upload
Applicants can upload their CV and submit the required information through a dedicated application flow.
Protected API Routes
Sensitive backend endpoints are secured through authentication middleware, authorization and request validation.
Application Status Tracking
Applicants can follow the progress of their submissions and view the current application status.
Technology stack
A complete full-stack architecture
The stack was selected to support responsive interfaces, secure APIs, scalable data and AI-assisted functionality.
Frontend
ReactTypeScriptViteTailwind CSSReact RouterAxios
Backend
Node.jsExpress.jsTypeScriptREST APIJWTbcrypt
Database
MongoDBMongoDB AtlasMongoose
AI & Deployment
OpenAI APIPrompt EngineeringGitHubPostmanRenderVercel
System architecture
How the application communicates
The frontend communicates with a protected backend API, which manages application data and AI service requests.
User Interface
React application
REST API
Node.js and Express
Database
MongoDB Atlas
AI Service
OpenAI integration
Authentication middleware validates user sessions before protected operations are completed. Role checks determine whether the user can access applicant, reviewer or manager functionality.
Engineering challenges
What required the most thought
The most valuable parts of the project were the decisions behind the implementation.
CHALLENGE 01
Designing role-based access
The platform needed separate interfaces and permissions for applicants, reviewers and managers. Protected routes and backend authorization rules ensure each role can only access the appropriate actions.
CHALLENGE 02
Structuring AI responses
AI responses had to be useful and predictable rather than unstructured paragraphs. The prompt and response flow were designed to produce structured candidate insights.
CHALLENGE 03
Handling authentication securely
Authentication required password hashing, token creation, protected API endpoints and secure frontend session handling.
CHALLENGE 04
Deploying multiple services
The frontend, backend and cloud database were deployed separately, requiring correct environment variables, CORS settings and production URLs.
Key learnings
What this project taught me
The project brought together several areas of software engineering in one end-to-end development process.
Designing a complete full-stack application from frontend to deployment
Creating reusable REST API endpoints and backend service layers
Building secure JWT authentication and role-based authorization
Integrating an AI service into a real product workflow
Working with file uploads and CV-related application data
Handling production deployment, environment variables and CORS
Structuring MongoDB collections for users and applications
Transforming an academic idea into a functioning digital product
Project gallery
A closer look at the product
Screenshots from the most important dashboards and user flows.
Explore more work
Interested in how I build complete digital products?
Explore my other projects or review the source code behind this application.