Backend engineer focused on building production-grade systems with Java & Spring Boot 3. I care about data consistency, concurrency control, failure handling, and clean domain modeling.

Live App · API Docs

Graduation project built with a cross-functional team. I led the backend.

• 30+ endpoint REST API with JWT/RBAC via Spring Security
• Integrated HuggingFace EfficientNetB3 Food-101 model via async pipeline (Spring @Async) — non-blocking photo analysis, frontend polls status endpoint
• Deployed on Render (Docker + MySQL); Flutter frontend on Firebase Hosting

Spring Boot 3 MySQL JWT HuggingFace Docker Flutter

Production-grade order management platform connecting companies with delivery providers.

• Multi-role RBAC via Spring Security @PreAuthorize; JWT + refresh token rotation with reuse detection
• Exactly-once state transitions via PostgreSQL unique constraint + pessimistic locking with @Version fallback
• Atomic per-IP rate limiting via Redis; dual-layer idempotency (DB + Redis) for safe network retries
• Immutable audit trail via @PreUpdate hook; full order state machine enforced at domain level

Spring Boot 3 PostgreSQL Redis Docker JUnit 5 Mockito

• 30+ endpoint API; decoupled order/payment flows via domain events (OrderCreatedEvent, PaymentCompletedEvent)
• Idempotent Stripe webhook (HMAC); two-phase inventory reservation + optimistic locking
• Redis cache-aside cut read latency ~40%; SQL double-entry bookkeeping for revenue reporting

Spring Boot 3 MySQL Redis Stripe Docker

Live App

• Backend enforces 7 hard + 5 soft scheduling constraints with async generation and status polling
• JWT + role-based access (ADMIN / SCHEDULER / INSTRUCTOR); PDF/Excel/PNG export
• React/TypeScript SPA with 20+ pages; CI/CD via GitHub Actions; deployed on Vercel

Spring Boot 3 MySQL React TypeScript Docker GitHub Actions

• Redis INCR + Base62 distributed ID generation with zero collisions; AtomicLong fallback on Redis failure
• Deterministic sharding (hashCode % N) — no routing table; atomic per-IP rate limiting via Redis
• Raw JDBC (no ORM) for explicit shard control

Spring Boot 3 MySQL Redis Raw JDBC Docker

• Pessimistic locking prevents concurrent double-bookings
• Interval overlap detection (start/end conflict query) — prevents time collisions at SQL level
• JWT + RBAC; clean layered architecture

Spring Boot 3 MySQL JWT Docker

• JPA Specification for dynamic multi-field filtering (title, location, salary, experience)
• Controlled application lifecycle (PENDING → ACCEPTED / REJECTED) with duplicate prevention
• Role-scoped access: employers manage jobs, candidates manage applications

Spring Boot 3 MySQL JWT JPA Specification

Concurrency      →  Pessimistic locking, optimistic locking, atomic operations
Consistency      →  Idempotency keys, state machines, transaction boundaries  
Failure Handling →  Token reuse detection, retry safety, graceful degradation
Security         →  JWT rotation, RBAC, rate limiting, webhook verification
System Design    →  Sharding, caching strategies, distributed ID generation

📫 mahmoudyoussef29@gmail.com · LinkedIn · Open to remote opportunities