Farida Hassan Ali
I build and ship AI systems, computer vision, deep learning, and the full-stack products that run them. Comfortable across the whole pipeline, from research to production. Open to AI/ML roles and internships.
About
I'm finishing a Software and Information Technology Engineering degree at the Egyptian Chinese University (3.85 CGPA), with hands-on experience across machine learning, computer vision, and NLP built through Microsoft's ML Engineer track at the Digital Egypt Pioneers Initiative and AMIT's AI diploma (92%). My work spans published research, real-time systems, and full-stack products.
Experience
- Developed practical machine learning and computer vision solutions through real-world projects.
- Applied data preprocessing, model training, evaluation, and deployment across the full ML lifecycle.
- Worked on AI system design and implementation workflows end to end.
- Built and optimized deep learning models for computer vision applications.
Projects
A dual-encoder transformer combining GraphCodeBERT and VulBERTa to detect vulnerabilities in Python code and classify them by CWE category. Published as a research paper on the Microsoft Community Hub and featured on ECU's official page.
My role: owned tokenization, loss function design, and the full training run built the GraphCodeBERT tokenization cache, co-developed checkpointing and validation, and implemented supervised contrastive loss and R-Drop KL divergence regularization.
A full-stack scheduling platform with a custom constraint-based solver that runs greedy assignment, conflict repair, and simulated annealing to minimize campus days and maximize schedule compactness across 12 coordinator-facing pages.
My role: built and maintained the reusable component architecture, contributed across coordinator and staff-facing pages, and enforced the navy/teal design system across the frontend.
Graduation project at DEPI, a real-time attendance built on YOLO face detection and ArcFace identity embeddings, with an anti-spoofing layer (texture analysis, motion, liveness) to block proxy check-ins, plus an admin dashboard with live monitoring and automated reports.
My role: cleaned face datasets and generated embeddings for identity matching, built the FastAPI backend, and implemented anti-spoofing and liveness detection.
Transfer learning with EfficientNet-B3 to sort waste into plastic, paper, glass, and metal for automated recycling.
Custom-trained YOLOv11 + OpenCV webcam app with a Flask interface, built for entry-gate safety scenarios.
K-Means customer segmentation, plus a Decision Tree / Logistic Regression model for house voting behavior.
Skills
Certifications
Statistics & linear algebra, Python for data science, ML, deep learning & transfer learning, NLP, computer vision, Azure AI Fundamentals & Azure AI Engineer Associate, MLOps, and prompt engineering.
Python, machine learning, deep learning, computer vision, NLP, data science, and MLOps.
Got something worth building?
I'm one message away. Tell me what you're working on. I read everything and reply fast.