AI Engineer specializing in LLMs, Transformer architectures, and production ML systems
Cairo, Egypt
I build production-grade AI systems with a focus on Large Language Models and deep learning. From implementing GPT architectures from scratch to deploying RAG pipelines that handle thousands of documents, I'm passionate about turning complex AI research into practical, scalable solutions.
Currently sharing what I learn through technical deep-dives on Medium and always excited to collaborate on challenging ML projects.
|
Built a 164M parameter GPT model implementing the complete Transformer decoder architecture in PyTorch. Includes Multi-Head Self-Attention, Causal Masking, and Sinusoidal Positional Encoding. Tech Stack:
|
Production RAG system for document Q&A with automatic source attribution. Achieved 15% retrieval precision improvement and handles 10,000+ embeddings with sub-second latency. Tech Stack:
|
|
ML ensemble model achieving 92.9% accuracy on 10,000+ athlete records. Deployed as Flask API with interactive dashboard for real-time risk assessment. Tech Stack:
|
Check out my repositories for more ML experiments, NLP projects, and AI implementations. |
LLM Fine-tuning (LoRA, PEFT) • RAG Systems • Vector Databases • Model Deployment • Transformer Architectures
Building: Advanced RAG architectures and LLM-powered applications
Learning: Scaling transformer models and optimizing inference pipelines
Writing: Technical deep-dives on GPT architecture,