At Turing, I contribute to training and evaluating a frontier coding-focused LLM, creating high-quality datasets across code execution, function calling, web browsing, and full-stack development.
I design evaluation frameworks for correctness, clarity, and reasoning quality, and I helped shape early tool-use, plugin integration, and autonomous workflow capabilities. I also lead and mentor pods of 20–30+ AI trainers while supporting coding tasks including unit testing, debugging, and SPA development.
For the World Bank, I build data-processing pipelines for large-scale economic analysis using Python, SQL, Pandas, Polars, Dask, and Azure Databricks. I optimized ETL processing from weeks to 4–5 hours and integrated web-scraping and ingestion workflows for downstream analysis.
My background also includes delivering 40+ freelance software projects, managing telecom and financial-services initiatives, and supporting telecom QoS equipment across Egypt.
