CV
Omar Alsaqa's Curriculum Vitae.
Contact Information
| Name | Omar Alsaqa |
| Professional Title | Machine Learning Engineer | Research Scientist |
| o.alsaqa@gmail.com | |
| Phone | +1 416-399-0632 |
| Location | Waterloo, Ontario |
Professional Summary
Research-driven Machine Learning Engineer with 4+ years of experience across industry and academia, specializing in Computer Vision, Graph Neural Networks (GNNs), and Generative AI. Proven track record of developing production-grade AI pipelines (MLOps) and publishing at top-tier venues including NeurIPS workshops and IEEE conferences.
Experience
-
2026 - 2026 Toronto, Canada
Research Assistant
Toronto Metropolitan University (Mitacs Accelerate Internship, with FenderAI)
- Supervisor: Dr. Rasha Kashef.
- Department of Electrical, Computer, & Biomedical Engineering.
- Developed an optimized OCR system for an automated verification pipeline, with more than 3x speedup.
-
2023 - 2025 Remote
Machine Learning Engineer
LightRing
- Developed a Highway vehicle speed estimation and road analysis system (83% accuracy in real-time).
- Automated the FTTH Design Process using GNNs with Setics Sttar, reducing manual design time by 70% (delivered as Windows App).
- Engineered a Multi-Agent assistant bot for the World Bank using LangGraph and Gemini API; implemented intelligent agent routing.
- Fine-tuned Whisper for Saudi dialect speech recognition, improving WER by 3%.
-
2022 - 2024 Remote
Research Fellow
Fatima Fellowship
- Developed BLISS (Bandit Sampling) for GNNs using DGL, improving training speed up to 15% and accuracy up to 18%.
- Work accepted at the MusIML workshop at NeurIPS 2025.
-
2021 - 2024 Egypt
Data Scientist
Orange Innovation Egypt
- Built MLOps Pipelines using MLflow, DVC, and Docker with real-time monitoring via Prometheus and Grafana.
- Automated ticket routing for France Business Operations (76% accuracy), reducing latency by 60%.
- Researched transformer-based Multiple Object Tracking (MOT) with 3D shape prediction.
- Deployed computer vision systems for plant disease detection and industrial OCR tracking (dot-matrix font).
-
2020 - 2020 Egypt
Computer Vision Intern
Intixel
- Architected a segmentation system for breast cancer micro-calcification using U-Net and GANs on DDSM-CBIS dataset.
-
2019 - 2020 Egypt
Deep Learning Intern
Valeo
- Developed a car command system and monitoring system for driver distraction.
- 2nd Place winner of the 2020 Dell Technologies Envision the Future Competition.
Education
-
2024 - 2026 Waterloo, Canada
Master's (Thesis based)
Wilfrid Laurier University
Applied Computing
- Thesis: Developed a novel optimized hierarchical Graph Neural Network for vision tasks, by designing a pure GNN layer that emulates depthwise separable convolutions.
- Supervisor: Prof. Emad A. Mohammed.
- Relevant Coursework: Advanced Parallel Programming (CP631), Advanced Machine Learning (CP640), Practical Algorithm Design (CP600), Applied Cryptography (CP614).
- Teaching Assistantships: Artificial Intelligence (CP468), Data Structures (CP264).
-
2015 - 2020 Mansoura, Egypt
Bachelor's
Mansoura University
Computer and Systems Engineering
Publications
-
GeoViG: Geometry-Aware Graph Reasoning for Mobile Vision Tasks in Natural and Medical Images
Accepted, IEEE EMBC 2026.
-
BLISS, Bandit Layer Importance Sampling Strategy for Efficient Training of Graph Neural Networks
NeurIPS 2025, MusIML Workshop.
-
Multimodal System for Driver Distraction Detection and Elimination
IEEE Access, 2022.
Honors and Awards
-
2026 Award for Outstanding Work at the Graduate Level
Wilfrid Laurier University
Selected to receive the award at the Spring 2026 convocation ceremony.
Projects
-
Parallel Classical vs. Graph-Based Image Denoising
CUDA kernels for up to 25x GPU speedup compared to single-threaded CPU baselines.
-
Neural Machine Translation (Google Trax)
Reproduced transformer performance; original PR contributor to Google Trax.
-
Multi-Agent LLM Automation Pipeline
Modular graph-based system using LangGraph, Gemini, and Jina AI.
Skills
Certificates
- Deep Learning Specialization - Coursera
- Machine Learning with Graphs (CS224W) - Stanford
- NLP with Deep Learning (CS224n) - Stanford
- Computer Vision (CS231n) - Stanford