Sarah Z Solieman

I am a data scientist focused on cybersecurity, MLOps, and production model deployment. I specialize in building scalable systems that mitigate financial and security risks and protect users from fraud and malicious activity.

I architect and launch cloud-based ML pipelines with automation for ingestion, hyperparameter tuning, validation, and deployment, and I have led cross-functional engineering teams to productionize models. I developed an XGBoost detector that achieved a 2.3% false positive rate at scale and hold a patent for a malicious URL classification platform.

I build tooling for model interpretability and reliability, including automated feature evaluation that halved feature sets and a Population Stability Index data-drift monitoring system to trigger proactive retraining. I also designed OCR validation pipelines and rapid clustering + ML blocking mechanisms that improved incident response and earned internal recognition.

I founded and led an innovation working group that combined NLP, network analysis, and social science to surface platform risks and coordinated manipulation. I bring a collaborative, impact-driven approach and enjoy mentoring teams to deliver defensible, production-grade ML solutions.