ML Engineer – VLM/VLA at MediCovate (2025-01 – Present)
AI Surgical Training
- Fine-tuned Qwen-VL on surgical datasets; designed continual pretraining strategy for domain adaptation.
- Architected multimodal pipeline: RGB video + instrument tracking + temporal motion for real-time skill scoring (SPARC, Normalized Jerk, Economy Index).
- Investigated SigLIP encoder internals; ran inference on NVIDIA VLA models for surgical scene grounding.
Founder & AI Engineer at Seher Al Hekayat (2026-01 – Present)
Arabic Personalized Children's Books
- Built end-to-end Arabic story generation pipeline (DeepSeek R1 + Seedream); fine-tuned Qwen-Image-Edit with LoRA for character consistency across multi-page books.
- Engineered RTL-aware PDF/PPTX export with Arabic font shaping and queue-based async job architecture.
Freelance at snabil tech (2025-01 – Present)
- Conducted market and supply chain studies
- Work on developing solutions
AI Engineer at AuraVisor (2025-01 – Present)
AI Skincare Advisor App
- Built bilingual (AR/EN) AI skincare advisor;
- Developed Expo React Native frontend with Laravel backend; implemented PostHog analytics and automated trend analysis pipeline.
Computer Vision & ML Engineer at Mind Cloud Engineering Team (2023-09 – 2024-08)
- Built and fine-tuned CV models for text detection, segmentation, and OCR optimization using PyTorch and TensorFlow.
