🏥 Smart Healthcare System — Backend

A production-ready, secure, and scalable RESTful API for a Smart Healthcare Appointment & Records System, built with Java 17 + Spring Boot 3.5. Implements real-world engineering practices including JWT-based auth with Redis-backed token revocation, role-based access control, centralized exception handling, request validation, OAuth2 social login, Razorpay payment gateway integration, billing management, Cloudinary-backed cloud file storage, server-side pagination, and automated API documentation.

Frontend Repository: 🔗 HealthCare-Frontend — React 19 | Vite | Redux Toolkit | Tailwind CSS | Razorpay Checkout

---

🎥 Video Walkthroughs (Explained by Me)

I've recorded short walkthroughs breaking down some of the trickier features and bugs in this project — not just showing that it works, but explaining the reasoning behind the implementation.

#	Topic	Link
1	🔔 Notification Feature — Overview	▶ Watch
2	🧩 Notification Feature — Service & Repository Layer	▶ Watch
3	✅ In-App Notification Feature — Completed Walkthrough	▶ Watch
4	🐞 JWT Authentication Bug Fix — Root Cause & Resolution	▶ Watch
5	✅ Bean Validation & Global Exception Handler	▶ Watch
6	📧 OTP-Based Registration Flow	▶ Watch
7	🔑 OTP Email-Based Password Reset	▶ Watch

💡 These videos are meant to give reviewers a look into my thought process — how I debug, design, and reason through real backend problems, not just the final code.

---

✨ Key Highlights (What Makes This Stand Out)

Feature	Details
🔐 JWT Auth + Role-Based Access	Stateless authentication with role-scoped endpoints (ADMIN / DOCTOR / PATIENT)
🚪 Redis-Backed Token Revocation	True server-side logout for stateless JWTs — a blacklisted token is rejected at the filter level even before its natural expiry, with the Redis key TTL matching the token's remaining lifetime exactly
⚡ Redis-Backed OTP Store	Email verification OTPs live in Redis with native key expiry (SETEX) instead of a manually-expired MySQL table — no cleanup job, no stale rows
📧 Email OTP Verification	6-digit OTP sent via email before registration; 10-minute expiry, single-use, auto-cleared on resend
🌐 Google OAuth2 Social Login	Patients and doctors can sign in with Google via Spring OAuth2 client
💳 Razorpay Payment Gateway	Real, verified online payments for appointment billing — UPI, Cards & Netbanking, with server-side signature verification
☁️ Cloudinary Cloud Media Storage	Profile pictures uploaded via multipart requests are validated, streamed, and persisted to Cloudinary — no local disk dependency, fully production-portable
📄 Server-Side Pagination	Every major list endpoint (Admin's Doctors/Patients/Billing, Doctor's Appointments/Patients, Patient's Doctors/Appointments/Prescriptions) accepts page/size query params and returns a Spring Data Page<T> instead of a full unbounded list
🛡️ Global Exception Handler	@RestControllerAdvice catches all exceptions — validation, auth, not-found, duplicates — and returns consistent JSON error responses with timestamp
✅ Bean Validation	@Valid + Jakarta Validation annotations (@NotNull, @NotBlank, @Email, @Digits) on all request DTOs
🩺 Doctor Approval Workflow	Doctors register but are locked out until an Admin approves their account
🔑 Password Reset Flow	Forgot-password → token generation → reset-password via secure token

---

🏗️ Auditing & Persistence Infrastructure

To maintain professional-grade data traceability, the system implements JPA Auditing.

• Automatic Metadata: Every core entity automatically records when it was created/modified and who performed the action.
• Traceability: Integrated with Spring Security to capture the currently logged-in user via AuditorAware.
• Implementation: Utilizes @MappedSuperclass with BaseAuditEntity to eliminate boilerplate, ensuring consistent created_at, updated_at, created_by, and updated_by fields across the entire database schema.

---

🏛️ Architecture

Classic 3-tier layered architecture, with Redis sitting alongside MySQL as a second, purpose-built data store for short-lived state:

Controller (REST API)  →  Service (Business Logic)  →  Repository (JPA / MySQL)
                                    │
                                    └──▶  RedisTemplate  →  Redis  (OTPs, JWT blacklist)

The codebase is organized by domain modules (feature-based packaging), not by layer — keeping related code co-located and the project scalable.

com.ankit.HealthCare_Backend/
├── appointment/          # Appointment booking, status updates, paginated repository queries
├── authentication/       # JWT + Redis blacklist, Redis-backed OTP, OAuth2, Security config
├── billing/              # Billing records, payment, revenue stats, Razorpay integration, paginated billing list
├── communication/        # Contact Us feature
├── core/                 # Shared enums (AppointmentStatus, BillingStatus), Role entity, RedisConfig
├── Exception/            # GlobalExceptionHandler + custom exceptions
├── filemanagement/       # Profile picture upload/retrieval, Cloudinary integration
├── Notification/         # Notification entity & repository
├── prescription/         # Doctor pre