Sarayah Wims

I’m an AI Evaluator and Data Annotator who helps teams ship safer, more accurate models by applying structured critical thinking to real-world outputs. At Handshake AI — Project Ohm, I evaluated AI-generated technical responses for accuracy and risk across multiple software systems, delivering real-time quality decisions.

I also thrive in data-heavy, deadline-driven environments. I’ve managed concurrent evaluation tasks independently while maintaining precision across high-volume annotation workflows, and I’ve contributed feedback that improved downstream model quality.

Beyond evaluation, I bring strong data organization and research discipline. As a Research Assistant in the Women In Natural Sciences Program, I entered, organized, and validated 20 years of multi-cohort records using Microsoft Excel and Microsoft 365, identifying and resolving inconsistencies to keep documentation audit-ready.

I’m equally comfortable building and communicating. My projects include RiskPulse (Python, Flask, Claude API, HTML/JS) for real-time AI risk monitoring with escalation logic, and Coding Our Air for CO2 trend analysis from EPA datasets using Python and Jupyter. I’m also building toward Google Cloud Generative AI leadership (GAIL) while staying grounded in responsible AI, clear communication, and continuous learning.