DevTrack is a full-stack platform that helps developers track their coding progress, manage projects, analyze resumes with AI, and measure career readiness — all in one unified workspace.
It aggregates activity from LeetCode, Codeforces, CodeChef, and GitHub, then uses AI to generate personalized learning roadmaps, career insights, and readiness scores.
| Feature | Description |
|---|---|
| DSA Progress Tracking | Unified view of problems solved across LeetCode, Codeforces, and CodeChef with heatmaps, topic analytics, and contest history |
| Platform Sync Engine | Automated background sync from 4 coding platforms using scheduled BullMQ workers |
| AI Resume Intelligence | Upload PDF/DOCX resumes for ATS scoring, claim validation, credibility analysis, and AI-powered improvement suggestions |
| Career Readiness Engine | Skill gap analysis, personalized roadmaps, evolution timelines, and an AI copilot for career guidance |
| Project Management | Track personal projects with tasks, velocity metrics, and GitHub repository integration |
| Focus Mode | Daily missions, Pomodoro timers, and persistent task lists to maximize productivity |
| Gamification | XP system, levels, streaks with freeze protection, missions, and daily challenges |
| Real-time Updates | Server-Sent Events (SSE) for live dashboard notifications and activity feeds |
| Admin Dashboard | Queue monitoring, dead-letter queue replay, trust scores, and AI audit logs |
┌─────────────────────────────────────────────────────────────────────┐
│ CLIENT (React 19 SPA) │
│ Zustand State • TanStack Query • SSE Client • Clerk Auth │
└────────────────────────────────┬────────────────────────────────────┘
│ REST + SSE
┌────────────────────────────────▼────────────────────────────────────┐
│ API SERVER (Express 5) │
│ 23 Domain Modules • Zod Validation • Rate Limiting • RBAC │
└───────┬───────────────────────┬──────────────────────┬──────────────┘
│ │ │
┌───────▼───────┐ ┌────────▼────────┐ ┌───────▼────────┐
│ MongoDB 7 │ │ Redis 7 │ │ BullMQ Workers │
│ 60+ Models │ │ Cache + PubSub │ │ 6 Worker Types │
└───────────────┘ └─────────────────┘ └────────────────┘
│
┌───────▼────────┐
│ External APIs │
│ LeetCode, CF, │
│ GitHub, OpenAI, │
│ Gemini │
└────────────────┘
Backend is modular with 23 domain modules following a clean service-controller-routes pattern. Each module has its own validation layer (Zod), business logic, and route definitions.
Workers handle platform sync, XP processing, resume intelligence (parsing, ATS scoring, embeddings), and system maintenance — all decoupled from the HTTP thread for low API latency.
- React 19 with Vite 8 for fast HMR and optimized builds
- TypeScript for full type safety across the application
- Tailwind CSS 4 with a custom design system and motion tokens
- Zustand for client-side state management with selective persistence
- TanStack Query for server state, caching, and optimistic updates
- Framer Motion for polished animations and transitions
- Clerk for authentication (OAuth, magic links, session management)
- Express 5 with async middleware and structured error handling
- MongoDB 7 with Mongoose ODM (60+ schemas, indexed queries)
- Redis 7 for caching, BullMQ job queues, and SSE pub/sub
- BullMQ for reliable background job processing with DLQ support
- Zod for runtime request validation on all endpoints
- Clerk Express SDK for JWT verification and user sync
- OpenAI GPT for resume analysis, coaching, and career intelligence
- Google Gemini for embeddings, semantic search, and fallback generation
- Custom credibility scoring algorithms and ATS simulation engines
- Docker Compose for local development and production deployment
- GitHub Actions CI/CD with automated testing and deploy pipelines
- Railway + Vercel for production hosting (API + static SPA)
- Split-process architecture: API and workers scale independently
DevTrack/
├── backend/ # Express 5 API + BullMQ Workers
│ ├── src/
│ │ ├── config/ # Environment and platform configuration
│ │ ├── db/models/ # 60+ Mongoose schemas
│ │ ├── middleware/ # Auth, rate limiting, error handling
│ │ ├── modules/ # 23 domain modules (routes/controller/service/validation)
│ │ ├── shared/ # Redis, SSE, queue utilities
│ │ └── workers/ # Background job processors
│ └── tests/ # Unit, integration, and chaos tests
│
├── frontend/ # React 19 SPA
│ ├── src/
│ │ ├── components/ # Reusable UI primitives and layouts
│ │ ├── features/ # Domain feature modules (15+)
│ │ ├── pages/ # Route-level page components
│ │ ├── services/ # Axios API client layer (16 modules)
│ │ └── store/ # Zustand state stores (14 stores)
│ └── e2e/ # Playwright end-to-end tests
│
├── docker-compose.yml # Full-stack development environment
├── docker-compose.prod.yml # Production split topology
└── .github/workflows/ # CI/CD pipelines
- Node.js >= 18
- Docker Desktop (for MongoDB + Redis)
# Clone the repository
git clone https://github.com/vbtgongithub/DevTrack.git
cd DevTrack
# Install all dependencies (backend + frontend)
npm run install-all
# Configure environment variables
cp backend/.env.example backend/.env
cp frontend/.env.example frontend/.env
# Start infrastructure (MongoDB + Redis)
npm run dev:infra
# Run the full application
npm run dev:all| Service | URL |
|---|---|
| Frontend | http://localhost:5173 |
| Backend API | http://localhost:3001 |
| API Docs | http://localhost:3001/health |
RESTful API mounted at /api/v1 with 23 route modules:
- Authentication — Clerk-based JWT with automatic user sync
- DSA & Platforms — Problem tracking, contest history, cross-platform analytics
- Resume Intelligence — Upload, parse, ATS score, credibility check, export
- Readiness — Skill gaps, roadmaps, evolution timeline, AI copilot
- Projects — CRUD with tasks, GitHub sync, velocity metrics
- Gamification — XP, streaks, missions, daily challenges, leaderboard
- Real-time — SSE event streams with Redis pub/sub backbone
- Operations — Queue health, DLQ management, system metrics
All endpoints use Zod schema validation, structured error responses, and rate limiting.
# Backend unit tests
npm test
# Backend integration tests (requires running infrastructure)
cd backend && npm run test:integration
# Frontend E2E tests (Playwright)
cd frontend && npm run test:e2eCI pipeline runs TypeScript type checking, linting, unit tests, integration tests, and Docker build verification on every pull request.
The application supports multiple deployment strategies:
- Docker Compose — Single command for the full stack
- Split Production — Independent scaling of API server and worker processes
- Cloud Native — Vercel (frontend) + Railway/Render (backend) with GitHub Actions automation
- 60+ MongoDB schemas with compound indexes and optimized query patterns
- 6 background worker types processing jobs asynchronously via BullMQ
- SSE real-time layer with Redis pub/sub for multi-instance broadcast
- Dead-letter queue with replay capability for fault tolerance
- Graceful AI degradation — platform works fully without AI keys in dev mode
- 14 Zustand stores with selective hydration and persistence
- Chaos testing suite — SSE reconnect storms, refresh floods, DLQ resilience
Private — All rights reserved.