Do you work with businesses in the UAE, Saudi Arabia, or other GCC countries?

Yes. Based in Lebanon, working remotely with clients worldwide — including the UAE, Saudi Arabia, Qatar, Kuwait, Bahrain, and Oman. Everything is scoped, delivered, and supported remotely.

Do you build custom websites and web applications?

Yes — end-to-end product builds: REST APIs, relational data models, and React frontends deployed on cloud infrastructure you can actually operate. Not a template, a custom build.

Can you build an AI agent or automation for my business?

Yes. Single-task agents (support triage, lead qualification, booking) wired into your CRM, WhatsApp, or site, built on n8n and LLM integration, or full multi-agent systems for larger workflows.

Do you work on computer vision or object detection projects?

Yes — training, optimizing, and deploying object detection and recognition models (YOLO, PyTorch) into a working product, not left as a notebook demo.

Can you run a security audit or penetration test on my application?

Yes. Backends are built with injection, XSS, and DoS mitigations from the first commit, and I maintain CAI, a dedicated autonomous penetration-testing framework covering web, API, cloud, and AI/ML attack surfaces.

Do you support Arabic and English?

Yes — EN/AR localization is available where scoped, relevant for GCC and wider Arabic-speaking audiences.

What does a project cost?

Pricing depends on the scope, integrations and support you need. Smaller, focused projects are welcome. Describe your project and I’ll suggest a practical scope and quote before work starts.

Describe your project or the change you need through the contact form below. We can discuss fit, scope and a quote before work starts.

Full Stack — AI — Security

Tyre, Lebanon

Custom software and AI automation for business operations.

I build web applications, workflow automations and AI agents for teams in Lebanon and the GCC. Bring a process your team handles by hand; we can scope the integrations, human handoff and a practical first release.

Discuss your workflow See project work

React, Django & APIs: Web apps
Agents, retrieval & computer vision: AI systems
Workflows & integrations: Automation
Lebanon & GCC delivery: Remote

What I build

Start with a workflow or application your business needs. Based in Lebanon, I work remotely with teams across the GCC and worldwide.

Workflow automation & AI agents

Connect enquiry intake, support triage, booking and reporting to the tools your team uses. Scope one workflow, its integrations and the points where a person stays in control.

n8n
LLM Integration
API Integration

Custom business software & web applications

Build a portal, internal dashboard or web application around your users and business rules. Define the data, permissions, integrations and handover alongside the features.

Python
Django REST
React
PostgreSQL
Azure

AI/ML engineering & integrations

Integrate computer vision, retrieval or language-model features into an application. Agree the evaluation approach, operating constraints and deployment requirements before expanding the scope.

YOLO
PyTorch
RAG Pipelines
LangChain

Application security & backend engineering

Review a defined web, API or cloud attack surface and turn findings into prioritized fixes. Security audit scope, testing authorization and reporting are agreed before work starts.

API Security
Penetration Testing
CAI Framework

Choose a starting point

Workflow automation →
Connect repeatable steps across your existing tools, with rules and clear exception handling.
Custom software development →
Build or improve a portal, dashboard or application around your users and business rules.
AI agent development →
Interpret requests or prepare tool actions within a bounded workflow, with human approval where needed.
RAG knowledge assistants →
Find and explain information from a maintained document collection, with useful source references.
Application security & penetration testing →
Review a defined web, API or cloud attack surface and turn findings into prioritized, actionable fixes.

Not sure which fits? Use the project brief checklist →

Pricing and engagement options are on the hire page.

Profile

Independent full-stack developer and AI engineer, based in Lebanon, working remotely with teams across the UAE, Saudi Arabia, Qatar, and beyond.
I design and build the custom software, workflow automation, and AI agents that solve a specific business process — not a generic template or chatbot demo.
Every project in this portfolio is my own build: personal research, academic work, or independent development. I'm upfront about which is which, and about what's been tested versus what hasn't.
Background in computer vision, cloud architecture (Azure), security-minded engineering, and AI-agent orchestration — applied to real, scoped problems, not kept as abstract skills.

Core competencies

Full Stack Development
AI/ML Engineering
Cloud Architecture (Azure)
Computer Vision & YOLO
Security & Penetration Testing
API Design & Integration
Database Optimization
Test Automation
System Architecture
LLM Integration & RAG
Workflow Automation
UX/UI Design (Figma)

Experience

Jun 2023 — Present
Full Stack Developer & AI Engineer
Freelance Full Stack Web Developer — Independent / Remote
- Develop full-stack applications connecting user interfaces, APIs, databases and cloud deployment.
- Build AI-powered applications, including YOLO ingredient detection and recommendation workflows.
- Implement input validation, access controls and application security measures appropriate to the project scope.
- Lead the complete development lifecycle from concept to deployment on Azure App Service, managing React frontend, Django backend, PostgreSQL databases, and ML components with automated testing and monitoring.
- Work on model inference and application performance, including quantization and caching approaches.
- Build and maintain multi-language support systems (English, Arabic, French) using Django's i18n framework, ensuring seamless user experience across diverse user demographics.
- Translate requirements into technical specifications, with delivery scope and acceptance checks agreed for the project.

Projects

Personal Research Initiative, 2024 — Present

CAI Security Framework

Lead Security Researcher & Developer

Personal security research framework for organizing test plans, tool execution and evidence across application, cloud and AI/ML attack surfaces.

Developed test cases for prompt injection, model extraction and adversarial attacks in a research setting.
Designed a container-based architecture with PostgreSQL and Elasticsearch, integrating security tools such as nmap, Metasploit and SQLMap.
Implemented evidence collection and SHA-256 hashing with reporting workflows. This research project does not establish regulatory compliance or certification.
Developed privilege escalation detection, lateral movement mapping, WAF evasion techniques, and AI-powered exploit chain generation.

Impact & differentiation

Why it matters —: The project explores a repeatable path from a security test plan to collected evidence and reporting.
Impact —: Structured evidence can support review of findings. Any business assessment still needs an agreed scope, authorization and interpretation by the responsible team.
Differentiator —: AI/ML test cases sit alongside application and infrastructure testing in the same personal research framework.

Tech deep-dive

Scalability —: Docker containers, PostgreSQL state and searchable evidence are the architecture described by the project. Deployment capacity requires separate validation.
Integration —: Tool execution and evidence collection are connected through a shared workflow rather than separate reports for each tool.
Security —: SHA-256 hashing helps check whether evidence has changed; it does not independently prove provenance, completeness or compliance.

Python
Docker
PostgreSQL
Elasticsearch
nmap
Metasploit
SQLMap

2023 — 2025

Food Ingredient Detection & Recommendation System

Research Developer — LIU University Master's Thesis

Master's thesis application integrating food recognition, recipe recommendations and a conversational interface.

Trained and optimized a YOLO object detection model for food-ingredient recognition as part of the thesis.
Architected a recipe recommendation engine using collaborative filtering and content-based algorithms, processing dietary restrictions in real time.
Built a conversational interface integrating Spoonacular, Edamam and TheMealDB APIs.
Implemented JWT authentication, SQL injection prevention, and input validation across all endpoints; deployed on React, Django REST, PostgreSQL, and Azure.

Impact & differentiation

Why it matters —: The thesis connects ingredient recognition to recipe recommendations and dietary preferences in one application.
Impact —: It demonstrates how a model can be integrated into an application workflow. It is academic work, without a claimed customer savings or support-resolution result.
Differentiator —: Computer vision, recommendation logic and external food-data APIs are combined in the same thesis project.

Tech deep-dive

Scalability —: Django REST + PostgreSQL on Azure App Service, with the ML inference path separated from the request/response cycle so the detection model can be scaled or swapped independently of the web tier.
Integration —: The recommendation engine blends collaborative filtering with content-based filtering rather than picking one — collaborative filtering alone cold-starts badly for new users, content-based alone misses cross-user patterns.
Security —: The implementation includes JWT authentication and input validation; these controls do not replace a deployment-specific security assessment.

YOLO
PyTorch
Django REST
React
PostgreSQL
Azure

Personal Productivity Project, 2024

AI-Powered Frontend Code Generator with Multi-Agent Reasoning

AI Agent Developer

Personal frontend code-generation project using CrewAI agent orchestration and example retrieval.

Implemented semantic few-shot learning with sentence-transformers for automatic example selection, integrating npm Registry, cdnjs, and GitHub APIs with 24-hour caching.
Built a CLI with resume input support (JSON/TXT/MD) for generating React/Vue code and animation examples. Generated output requires review, testing and accessibility checks.

Impact & differentiation

Why it matters —: Semantic example retrieval supports agent prompts using relevant examples from a shared collection.
Impact —: Caching registry and repository API responses is intended to reduce repeated requests; no benchmark or business time-saving result is claimed here.
Differentiator —: Resume-aware generation (JSON/TXT/MD upload) ties output style to the actual developer's stack instead of a generic component template.

Tech deep-dive

Scalability —: Separate agent roles are coordinated through CrewAI. Additional roles still require task-specific evaluation.
Integration —: sentence-transformers + FAISS handle example retrieval instead of hand-picked few-shot examples per prompt, so the example set can grow without blowing up prompt length.
Security —: Accessibility requirements can guide generation, but generated components still need keyboard, screen-reader and contrast verification. No automatic WCAG compliance is claimed.

Python
CrewAI
LangChain
Mistral AI
FAISS
Sentence Transformers

Independent Project, 2023

GPS Vehicle Tracking System

Full Stack Developer

Independent GPS tracking project connecting an Android application, Firebase and a web dashboard.

Read the project walkthrough →

Firebase
JavaScript
Android

Personal Productivity Project, 2024

Local AI Agent Toolkit

AI Systems Developer

Personal agent toolkit for development tasks including frontend, backend, testing and workflow automation.

Read the project walkthrough →

Python
ChromaDB
LangChain
RAG

Personal Productivity Project, 2024

N8N Workflow Analysis & Auto-Fix System

AI Agent Developer

Personal n8n workflow-analysis tool with schema, security, semantic and performance checks and supported fixes.

Read the project walkthrough →

n8n
Python
Schema Validation

Skills

Languages

Python
JavaScript
SQL
PHP
Java
HTML/CSS

AI & Machine Learning

YOLO
PyTorch
Computer Vision
Object Detection
Model Optimization
RAG Pipelines
LangChain
LLM Integration

Frameworks & Libraries

Django
Django REST Framework
React.js
Django ORM
Robot Framework

Databases

PostgreSQL
Elasticsearch
Firebase Realtime Database
SQL Optimization

Cloud & DevOps

Azure App Service
Azure AI Services
Firebase
Docker
CI/CD
API Integration

Security

Penetration Testing
SQL Injection Prevention
XSS Protection
DoS Mitigation
Compliance Frameworks

Testing & Automation

Robot Framework
Unit Testing
Integration Testing
Performance Testing

Design & UX

Figma
Multi-language UX (i18n)
Chatbot Interface Design
Responsive Design

Education

Oct 2023 — Jun 2025
Master of Science in Computer Engineering
LIU University — Tyre, Lebanon
Feb 2018 — Jun 2023
Bachelor of Science in Computer Engineering
LIU University — Tyre, Lebanon

Certifications & Training

Microsoft Azure Learning Paths

Azure Software Developer — Completed
Azure AI Engineer — Near Completion

LinkedIn Learning Certifications

Offensive Penetration Testing
Robot Framework Test Automation (Level 1 & 2)
Advanced Chatbot Development
Fine-Tuning LLMs for Cybersecurity
Figma for UX Design

Questions clients actually ask

Do you work with businesses in the UAE, Saudi Arabia, or other GCC countries?: Yes. Based in Lebanon, working remotely with clients worldwide — including the UAE, Saudi Arabia, Qatar, Kuwait, Bahrain, and Oman. Everything is scoped, delivered, and supported remotely.
Do you build custom websites and web applications?: Yes — end-to-end product builds: REST APIs, relational data models, and React frontends deployed on cloud infrastructure you can actually operate. Not a template, a custom build.
Can you build an AI agent or automation for my business?: Yes. Single-task agents (support triage, lead qualification, booking) wired into your CRM, WhatsApp, or site, built on n8n and LLM integration, or full multi-agent systems for larger workflows.
Do you work on computer vision or object detection projects?: Yes — training, optimizing, and deploying object detection and recognition models (YOLO, PyTorch) into a working product, not left as a notebook demo.
Can you run a security audit or penetration test on my application?: Yes. Backends are built with injection, XSS, and DoS mitigations from the first commit, and I maintain CAI, a dedicated autonomous penetration-testing framework covering web, API, cloud, and AI/ML attack surfaces.
Do you support Arabic and English?: Yes — EN/AR localization is available where scoped, relevant for GCC and wider Arabic-speaking audiences.
What does a project cost?: Pricing depends on the scope, integrations and support you need. Smaller, focused projects are welcome. Describe your project and I’ll suggest a practical scope and quote before work starts.
How do we start?: Describe your project or the change you need through the contact form below. We can discuss fit, scope and a quote before work starts.

Custom software and AI automation for business operations.

What I build

Workflow automation & AI agents

Custom business software & web applications

AI/ML engineering & integrations

Application security & backend engineering

Profile

Core competencies

Experience

Full Stack Developer & AI Engineer

Projects

CAI Security Framework

Food Ingredient Detection & Recommendation System

AI-Powered Frontend Code Generator with Multi-Agent Reasoning

GPS Vehicle Tracking System

Local AI Agent Toolkit

N8N Workflow Analysis & Auto-Fix System

Skills

Languages

AI & Machine Learning

Frameworks & Libraries

Databases

Cloud & DevOps

Security

Testing & Automation

Design & UX

Education

Certifications & Training

Questions clients actually ask

Let's build something worth shipping.