Eyad Harb

// 01. Introduction

About Me

I'm a Full Stack Developer and AI/NLP Engineer specializing in building intelligent, scalable applications that merge cutting-edge machine learning with modern web technologies.

With expertise in Python, TensorFlow, PyTorch, React, and Angular, I've delivered impactful projects including an accessible job platform serving 500+ listings and an Arabic fake review detector achieving 85% accuracy.

My work focuses on creating accessible, performant solutions deployed on AWS with Docker and CI/CD, always prioritizing user experience and technical excellence.

500+

Job Listings Managed

85%

ML Model Accuracy

Major Projects

100+

Bookings Processed

// 02. Portfolio

Featured Projects

Team Lead

EMPOWERABILITY: Jobs Without Limits (AbilityHire)

Yarmouk University | Feb 2025 – Jun 2025

Developed comprehensive job-seeking platform with accessible UI for blind users
Designed secure admin dashboard managing 500+ job listings with Flask and MySQL
Integrated Google Maps API for location-based job information, improving UX by 30%
Built database schema from scratch with optimized queries for real-time data retrieval
Ensured WCAG compliance with screen reader compatibility and keyboard navigation
Implemented voice control via Web Speech API

Flask MySQL HTML/CSS/JS Google Maps API Playwright E2E WCAG

ML Engineer

Arabic Fake Review Detection

Yarmouk University | Jan 2024 – Dec 2024

Fine-tuned AraBERT/CamelBERT on 10k+ Arabic reviews
Achieved ~85% accuracy and macro-F1 ≈ 0.83
Deployed model via FastAPI + Docker on AWS with p95 < 120ms
Implemented drift monitoring and comprehensive logging systems
Handled Arabic normalization, diacritics, dialect variation, and sarcasm/code-mix

PyTorch AraBERT CamelBERT FastAPI Docker AWS Python

Project Lead

Hotel Management System

Yarmouk University | 2023

Led full-cycle development of hotel booking system
Created booking & inventory logic with reporting dashboard
Simulated and processed 100+ reservations
Managed 3-member team with Agile planning
Reduced delivery delays by 20%

PHP MySQL JavaScript UML DFD Agile

NLP Sentiment Analysis API

2024

Built LSTM-based classifier with Flask API
Achieved ~0.82 macro-F1 score for real-time sentiment analysis
Compared baseline SVM with TF-IDF vs deep learning pipelines
Implemented RESTful API with efficient JSON response handling

TensorFlow LSTM Flask scikit-learn SVM TF-IDF Python

Ranked Information Retrieval System

2024

Built search prototype using TF-IDF/BM25 algorithms
Indexed over 5k+ documents with 25% relevance improvement
Prepared infrastructure for future RAG integration
Implemented hybrid retrieval mechanisms for enhanced accuracy

Python TF-IDF BM25 NLP RAG

IoT Smart House Simulation

Multimedia Project | July 2025

Built interactive animation using Animaker showcasing IoT devices
Simulated lighting, climate, irrigation, and security systems
Designed comprehensive storyboard and logic flow
Presented functional scenarios simulating user-device interactions

Animaker IoT Multimedia Design Storyboarding

// 03. Technical Stack

Skills & Technologies

Languages

Python
JavaScript
PHP
C++
C#
SQL

ML/AI Tools

TensorFlow, PyTorch, Keras
scikit-learn
HuggingFace Transformers
AraBERT, CamelBERT
NLP & Text Classification

Front-End

HTML, CSS, JavaScript
React.js
Angular
Bootstrap

Back-End

Flask, FastAPI
.NET Core
Node.js
RESTful APIs

Database

MySQL
Oracle DB
Database Design
ER Diagrams

Cloud & DevOps

AWS (EC2, S3, Lambda, RDS)
Docker
CI/CD Pipelines
Git/GitHub

Testing

Playwright (UI Automation)
PyTest, Jest
Unit Testing
Manual QA

Additional

Web Accessibility (WCAG)
API Integration
Agile/Scrum
UML, DFDs
System Design