Built by an engineer who researches, ships, and publishes.

I graduated with Honours in Computer Engineering from The American University in Cairo, where I developed a deep focus on Generative AI and Multimodal LLMs. My final-year thesis — sponsored by Microsoft — was accepted at IEEE CAI 2026, introducing a table-first guarded reasoning pipeline that achieved 93.96% accuracy on the ChartQA benchmark while outperforming models 10× its size.

Professionally, I have built production ETL systems for the National Bank of Egypt, engineered AI ingestion pipelines for the AUC Digital Library, and currently work with Outlier to improve the reasoning capabilities of frontier large language models. My work spans the full stack — from low-level signal processing and model fine-tuning to cloud infrastructure and user-facing products.

Zenith Labs exists because I believe elite software should not require a 50-person agency. I work with a small number of clients at a time, applying research-grade rigour to every system I build — and delivering work I would put my name on a paper for.