Ziad Ashraf

I design and build scalable data platforms, from ingestion and ETL to analytics, machine learning, deep learning, and generative AI deployment. My work spans data engineering, MLOps, BI, and ML/DL/LLM-based applications, creating systems that are reliable, automated, and ready for production. I prioritize clean code, modular architecture, and measurable impact.

Machine Learning & AI: I handle feature engineering, model development, and deployment. I work across regression, clustering, classification, forecasting, NLP, and LLMs using Scikit-learn, TensorFlow, PyTorch, and Hugging Face. I deploy models with FastAPI and Flask following MLOps best practices: versioning, monitoring, CI/CD, and automated pipelines.

Data Engineering: I design and orchestrate ETL/ELT pipelines using Airflow, dbt, Spark, and Airbyte across AWS, GCP, and Azure. I build real-time data systems with Kafka, Debezium, and Flink; implement data warehouse architectures; and ensure data quality with automated testing, observability, and DataOps methodologies.

Analytics & BI: I turn raw data into actionable insights using SQL, Python, and R. I build KPI-driven dashboards in Power BI, Tableau, Looker, and Superset, and conduct statistical analysis, A/B testing, cohort analysis, and marketing analytics to guide business decisions.

Whether it’s data orchestration, business intelligence, or predictive modeling, I focus on building reliable, scalable solutions that deliver real value.