Skip to content
BranchesTags

Repository files navigation

PlaceTrack AI

Enterprise-grade Placement Management & Student Readiness Platform for Engineering Colleges

Live Demo Next.js React TypeScript Express PostgreSQL Prisma Gemini AI Docker Vitest License API Health

🚀 Live Demo → frontend-umber-one-zuhj8ueccq.vercel.app

Quick Links

A full-stack placement management platform enabling students to track applications, analyze resumes with AI, practice aptitude tests, and prepare for interviews — while coordinators and admins manage drives, pipelines, and reports from a unified dashboard.

Screenshots

Dashboard Applications
Dashboard Applications
Opportunities Resume AI
Opportunities Resume AI
Aptitude Test Profile
Aptitude Profile
Interview Coach AI Feedback
Interview AI Feedback

Features

Student Portal

  • Readiness score ring computed from CGPA, aptitude, coding, communication, projects, internships, and mock counts
  • Job board with real-time eligibility check and suitability score per drive
  • AI-powered resume analysis via Google Gemini with heuristic fallback
  • Timed aptitude mock tests with server-side grading and section-wise feedback
  • AI interview coach generating role-specific questions with answer evaluation
  • In-app notifications for application status changes and interview schedules
  • 90 skills across 4 categories (Technical Core, Tools & Platforms, Core Engineering, Soft Skills) covering all 10 branches — CS, IT, AI/DS, Electronics, Electrical, Mechanical, Civil, Chemical, and more

Coordinator & Admin Portal

  • Create and manage placement drives with branch filters, CGPA caps, backlog limits, and deadlines
  • Progress student applications through a strict pipeline: APPLIED → SHORTLISTED → APTITUDE_CLEARED → TECHNICAL_ROUND → HR_ROUND → SELECTED / REJECTED
  • Schedule interviews and push notifications to candidates
  • Export full student registry and application data as CSV reports
  • Admin user management with audit trail logging

Architecture

graph TD
    Client[Next.js Frontend] <-->|JSON + JWT Auth| Express[Express.js Server]
    Express <-->|Prisma Client| DB[(PostgreSQL on Neon)]
    Express <-->|REST API| Gemini[Google Gemini AI]
    Express -->|Fallback| Heuristics[Regex Heuristic Parser]

Monorepo Structure

├── backend/
│   ├── prisma/
│   │   ├── schema.prisma          # Database schema
│   │   └── seed.ts                # 150-student seed script
│   ├── src/
│   │   ├── lib/                   # Prisma client & audit logger
│   │   ├── middleware/            # JWT auth & RBAC guards
│   │   ├── routes/                # REST endpoints by resource
│   │   ├── services/              # Eligibility, readiness & AI integrations
│   │   ├── types/                 # TypeScript annotations
│   │   └── server.ts              # Server entrypoint with auto-seed
│   └── tests/                     # Vitest unit test suite
│
├── frontend/
│   └── src/
│       ├── app/                   # Next.js App Router
│       ├── components/            # Dashboard, Resume AI, UI components
│       └── lib/                   # API wrappers, types, credentials
│
├── docker-compose.yml             # PostgreSQL container (local dev)
├── package.json                   # Monorepo workspaces config
└── .env.example                   # Environment variable template

Core Algorithms

Student Readiness Predictor

A backend scoring engine that maps academic and assessment metrics to a placement probability score:

Readiness Score = clamp(
  (CGPA / 10 × 22) +
  (Aptitude × 0.18) +
  (Coding × 0.22) +
  (Communication × 0.15) +
  (Projects × 4) +
  (Internships × 5) +
  (Mock Tests × 1.2) -
  (Active Backlogs × 7)
)
Range Status
>= 80 Placement Ready
65 – 79 Nearly Ready
< 65 Needs Focused Preparation

Placement Eligibility Engine

Hard constraints validated before a student can apply to any drive:

  • CGPA >= drive minimum
  • Active backlogs <= drive maximum
  • Branch within allowed departments
  • Graduation year matches drive cohort

Eligible applications receive a suitability score (50–100) based on academic margin (+15), skill keyword overlap (+30), and branch specificity (+15).

AI Resume Analyzer

Resume PDFs are parsed via pdf-parse into raw text, then sent to Google Gemini (gemini-2.5-flash) with a structured JSON prompt. Response fields: score, skills, sectionHits, suggestions, contactComplete. A regex heuristic fallback activates automatically if the API key is absent or rate-limited — the platform remains fully functional offline.

API Reference

Quick Overview

  • Backend Base URL: http://localhost:4000 (Local) / https://your-backend-url (Production)
  • Health Check: GET /health → returns { "status": "ok", "database": "connected" }
  • Authentication: All protected endpoints require Authorization: Bearer <jwt_token> header.

Endpoints Table

Method Endpoint Auth Description
GET /health Public Database connectivity health check
POST /api/auth/login Public Login and receive signed JWT
POST /api/auth/signup Public Register student or coordinator
GET /api/auth/me User Fetch profile and unread notifications
PATCH /api/auth/me/student Student Update profile and recalculate readiness
GET /api/auth/users Admin List all registered accounts
DELETE /api/auth/users/:id Admin Delete account and dependencies
GET /api/dashboard User Role-specific aggregated dashboard data
GET /api/drives User List drives with eligibility state per student
POST /api/drives Coordinator, Admin Create a placement drive
GET /api/applications User Query applications (owner-scoped for students)
POST /api/applications Student Submit eligibility-filtered application
PATCH /api/applications/:id/status Coordinator, Admin Advance application pipeline stage
POST /api/applications/:id/interview Coordinator, Admin Schedule interview and notify student
GET /api/tests User List all aptitude tests
GET /api/tests/:id User Fetch test details and questions
POST /api/tests/:id/submit Student Submit answers and record scored result
POST /api/ai/resume/text Student Analyze raw text resume
POST /api/ai/resume/upload Student Extract and analyze PDF resume
POST /api/ai/interview User Generate role-specific interview questions
POST /api/ai/interview/feedback User Evaluate answer with score and model response
GET /api/reports/applications.csv Coordinator, Admin Export application data as CSV
GET /api/reports/students.csv Coordinator, Admin Export student performance registry as CSV

Getting Started

Prerequisites

  • Node.js v20+
  • Docker Desktop (for local PostgreSQL)

Step 1 — Environment Setup

Copy .env.example templates in both backend/ and frontend/ directories:

# Copy backend environment configuration template
cp backend/.env.example backend/.env

# Copy frontend environment configuration template
cp frontend/.env.example frontend/.env.local

Backend Environment Variables (backend/.env.example)

Variable Description
PORT Express REST API server port (default: 4000)
DATABASE_URL PostgreSQL pooled connection string for Prisma Client
DIRECT_URL Direct PostgreSQL connection string for Prisma migrations
JWT_SECRET Secret string for signing and verifying JWT tokens
GEMINI_API_KEY Google Gemini API key for AI resume parsing & interview feedback
FRONTEND_URL Allowed CORS origin & email link base URL (default: http://localhost:3000)

Frontend Environment Variables (frontend/.env.example)

Variable Description
NEXT_PUBLIC_API_URL Backend REST API base URL (default: http://localhost:4000)

Step 2 — Start Database

docker compose up -d

Step 3 — Install, Migrate & Seed

# Install all workspace dependencies
npm install

# Push Prisma schema to database
npm run prisma:push -w backend

# Seed with 150 student profiles, companies, drives, and tests
npm run prisma:seed -w backend

Step 4 — Run Development Servers

npm run dev
Service URL
Frontend http://localhost:3000
Backend API http://localhost:4000
Health Check http://localhost:4000/health

Demo accounts are pre-seeded. Refer to backend/prisma/seed.ts for login credentials.

Testing

# Run backend unit tests
npm test

# TypeScript type check across all packages
npm run typecheck

Security

  • JWT authentication with HS256, 12-hour expiry
  • Role-based access control (STUDENT, COORDINATOR, ADMIN) enforced via middleware
  • Helmet for secure HTTP headers
  • Rate limiting: 300 requests / 60 seconds per IP on all /api/* routes
  • Zod schema validation on all request bodies before database calls
  • In-memory PDF processing via multer.memoryStorage() — no resume files written to disk
  • Cache-Control: no-store on all API responses

Database Schema

Model Purpose Key Indexes
User Auth entity with role Unique email
Student Academic profile [branch, graduationYear], [readinessScore]
Coordinator Staff department info Unique userId
Company Recruiter registry Unique name
PlacementDrive Job opening with constraints [status, deadline], [graduationYear]
Application Student-Drive relationship Unique [studentId, driveId], [status]
Interview Scheduled interview per application Unique applicationId
AptitudeTest Test blueprint —
Question MCQ per test —
TestResult Submission scores and metrics Unique [studentId, testId]
Notification In-app user alerts —
ActivityLog Admin audit trail [timestamp], [resource]
ResumeAnalysis Persisted AI scan results [userId, createdAt]

License

MIT © Sanket Chaudhari

About

Enterprise-grade placement management platform with AI resume analysis, readiness scoring, aptitude tests, and interview coaching for engineering colleges.

frontend-umber-one-zuhj8ueccq.vercel.app

Topics

dockerexpressjsfull-stackgemini-aigemini-apinextjsplacement-managementplacement-preparationpostgresqlprismareactrender-deploymentstudent-portaltypescriptvercel-deployment

Resources

Readme
MIT license

Code of conduct

Code of conduct

Contributing

Contributing
Activity

Stars

1 star

Watchers

0 watching

Forks

0 forks
Report repository

Releases

Packages

Contributors

Languages