Typing SVG
Subtitle
┌──────────────────────────────────────────────────────────────────┐
│ │
│ "I don't just write code. I design the systems │
│ that make code worth writing." │
│ │
│ 8h → 1h pipelines · 11,500+ users · 97% crash-free │
│ 14+ stores unified · 200% traffic growth │
│ │
└──────────────────────────────────────────────────────────────────┘
┌─────────────────────────────────────────────┐
│ ABDULRAHMAN AL-FAKIAH │
│ Full-Stack Systems Engineer │
└────────────────────┬────────────────────────┘
│
┌──────────────────────────┼──────────────────────────┐
│ │ │
┌──────────▼──────────┐ ┌──────────▼──────────┐ ┌──────────▼──────────┐
│ DATA ENGINEERING │ │ BACKEND PLATFORMS │ │ MOBILE & CLIENT │
│ │ │ │ │ │
│ Python · Prefect │ │ Node.js · TypeScript│ │ React Native │
│ CrawleeJS · ETL │ │ Laravel · Filament │ │ Expo · FlashList │
│ PostgreSQL │ │ REST · SDUI · RBAC │ │ Reanimated │
│ AWS S3 · Typesense │ │ Search Pipelines │ │ TanStack Query │
│ Stream Processing │ │ Audit & Admin Tools │ │ AI Chat Streaming │
│ Docker · Nginx │ │ Multi-tenant Auth │ │ RTL/i18n · CI/CD │
└──────────┬──────────┘ └──────────┬──────────┘ └──────────┬──────────┘
│ │ │
└──────────────────────────┼──────────────────────────┘
│
┌──────────▼──────────┐
│ LEADERSHIP │
│ │
│ Architecture Owner │
│ Technical Mentor │
│ PR Review Culture │
│ Agile Delivery │
│ Prod Monitoring │
└─────────────────────┘
Python Prefect Node.js PostgreSQL Typesense AWS S3 AWS EC2 Docker
🧠 GRAPH-BASED IDENTITY RESOLUTION ENGINE — Click to expand the madness
THE PROBLEM:
→ 14+ e-commerce stores
→ Same product, different names, specs, formats
→ No unified identity layer
→ Users couldn't compare prices across stores
MY SOLUTION:
┌──────────────┐ ┌──────────────┐ ┌──────────────┐
│ Store A │ │ Store B │ │ Store N │
│ "iPhone 15" │ │ "Apple IP15" │ │ "iP 15 Pro" │
└──────┬───────┘ └──────┬───────┘ └──────┬───────┘
│ │ │
└───────────────────┼───────────────────┘
│
┌────────▼────────┐
│ GRAPH ENGINE │
│ (PostgreSQL) │
│ │
│ Node = Product │
│ Edge = Match │
│ Weight = Score │
└────────┬────────┘
│
┌────────▼────────┐
│ CANONICAL ID │
│ 80% coverage │
│ → 200% traffic │
└─────────────────┘
RESULT: Unified product identities across 14+ stores.
80% stable ID coverage. 200% traffic increase.
Clustering features that users actually use.
⚡ 87.5% PIPELINE ACCELERATION — 8 hours to 1 hour. Yes, really.
BEFORE (8 HOURS):
┌─────────┐ ┌─────────┐ ┌─────────┐ ┌─────────┐
│ Scrape │────▶│ Parse │────▶│ Store │────▶│ Index │
│ (seq.) │ │ (sync) │ │ (batch) │ │ (slow) │
└─────────┘ └─────────┘ └─────────┘ └─────────┘
⏱️ 8h total — running overnight, praying nothing breaks
AFTER (1 HOUR):
┌─────────┐
│ Scrape │──┐ ┌─────────┐ ┌──────────┐ ┌─────────┐
│ (par.) │ ├─▶│ Parse │────▶│ FS Queue │────▶│ S3 + │
│ Scrape │ │ │ (stream)│ │ (stream) │ │ Index │
│ (par.) │──┘ └─────────┘ └──────────┘ └─────────┘
│ Scrape │
└─────────┘
⏱️ 1h total — parallelized, streamed, bulletproof
HOW:
→ Refactored CrawleeJS pipeline architecture from scratch
→ Introduced parallel execution across routes
→ Redesigned orchestration & routing logic
→ Built filesystem queue + stream processing layer
→ Result: 87.5% reduction. Data fresh by morning.
📱 RAKHYS MOBILE PLATFORM — 11,500+ USERS — The one on the App Store
┌──────────────────────────────────────────────────┐
│ RAKHYS MOBILE APP │
│ React Native · Expo · 11,500+ users │
├──────────────────────────────────────────────────┤
│ │
│ ✅ 97% crash-free rate (Firebase Crashlytics) │
│ ✅ 10% affiliate CTR increase │
│ ✅ 25% FPS improvement (FlashList + Reanimated) │
│ ✅ Full RTL / Arabic-first support │
│ ✅ AI chat with streaming interface │
│ ✅ Server-Driven UI — OTA layout updates │
│ ✅ v6.0 complete UI/UX redesign │
│ ✅ CI/CD pipeline for seamless releases │
│ │
│ 📲 Live on iOS & Android │
│ │
└──────────────────────────────────────────────────┘
🎛️ SERVER-DRIVEN UI FRAMEWORK — Deploy mobile updates without deploying
THE CONCEPT:
Non-engineers needed to change app layouts,
banners, and promotions — WITHOUT app releases.
THE ARCHITECTURE:
┌───────────────┐ ┌───────────────┐
│ ADMIN PANEL │────────▶│ SDUI CONFIG │
│ (Filament) │ │ (JSON/API) │
└───────────────┘ └───────┬───────┘
│
┌───────▼───────┐
│ MOBILE APP │
│ Renders UI │
│ from config │
│ │
│ No release │
│ needed ✨ │
└───────────────┘
IMPACT:
→ Marketing team ships layout changes same-day
→ Zero app store review delays for content updates
→ Operational autonomy for non-technical stakeholders
2024 ─────────────────────────────────────────────────────── PRESENT
│
├─ 🏢 NEXUMIND — Lead Full-Stack Engineer (Remote, Saudi Arabia)
│ └─ Rakhys ecosystem: data pipelines + backend + mobile
│ └─ Architecting systems serving 11,500+ users
│ └─ Python/Prefect orchestration, React Native, TypeScript
│
2022 ──────────────────────────────────────────────────────── 2024
│
├─ 🏢 HORIZONTAL — Backend Developer (Sana'a, Yemen)
│ └─ Laravel REST APIs, multi-tenant RBAC
│ └─ 20% API latency reduction
│
├─ 🏢 ARTECH SOFTWARE SOLUTIONS — Full-Stack Developer
│ └─ E-commerce platform (Laravel + Flutter)
│ └─ FCM push notifications, Firebase integration
│
2021 ──────────────────────────────────────────────────────── 2022
│
├─ 🔬 IUTT RESEARCH TEAM — Junior ML Engineer
│ └─ Masked face recognition (93% accuracy)
│ └─ VGG16/ResNet transfer learning
│ └─ Published in Springer ICESP 2023
│
2018 ──────────────────────────────────────────────────────── 2022
│
└─ 🎓 B.Sc. Computer Science
└─ International University of Technology Twintech
Publication
"Masked Face Recognition Using Transfer Learning Approaches"
4th International Conference on Electronics and Signal Processing (ICESP 2023)
Publisher: Springer, Cham
DOI: 10.1007/978-3-031-36670-3_3
Methodology: VGG16 / ResNet · Transfer Learning
Accuracy Achieved: 93%
🌐 LET'S CONNECT, Telegram
┌──────────────────────────────────────────────────────────┐
│ │
│ I build systems that move data, serve users, │
│ and scale without drama. │
│ │
│ If you're working on something ambitious — │
│ let's talk. │
│ │
│ 📧 abdulrahman1ye@gmail.com │
│ 📍 Sana'a, Yemen │
│ 📞 │
│ │
└──────────────────────────────────────────────────────────┘
Profile Views
$ echo "Thanks for scrolling this far. You're clearly someone with taste."
⚡ Last updated: 2025 · Built with obsessive attention to detail