Tarek Masryo
Building decision-ready ML & GenAI systems that improve operations, reduce risk, and accelerate workflows.
Building decision-ready ML & GenAI systems that improve operations, reduce risk, and accelerate workflows.
Production-ready ML & GenAI systems, shipped end to end
Hi, I'm Tarek Masryo — an AI/ML engineer who builds production-ready ML and GenAI systems that improve decisions, reduce operational risk, and accelerate workflows.
Python-first with a strong CS foundation (C/C++), I deliver end-to-end products: data pipelines, robust models, and decision-support dashboards/APIs that turn analytics into action, built to be maintainable, reproducible, and easy to run beyond notebooks.
My work spans NLP and Computer Vision, alongside grounded LLM applications (retrieval + tool use) designed for real-world constraints and evaluated for quality, latency, and cost.
I publish and maintain projects on GitHub, Kaggle, and Hugging Face — showcasing practical systems thinking from data and modeling through deployment and monitoring.
Core work areas with a production mindset
Clean data → features → training → evaluation → inference-ready packaging with reproducibility at the core.
KPI-first UX, drill-down analytics, and decision-focused insights with Streamlit and interactive visualizations.
Structured outputs, validation, lightweight safety, and evaluation-friendly logs for production systems.
Reproducibility, CI-friendly workflows, artifact management, and deployable services with FastAPI & Docker.
Selected work across different domains
What I use to build, ship, and operate production ML/GenAI systems
Open to AI/ML Engineer roles focused on production ML, MLOps, and reliable GenAI/RAG.