onesmas kamau

I’m an analytical, detail-oriented, systems-driven professional with a strong background in economics and advanced data analysis. I focus on optimizing multilingual data curation, filtering, and model evaluation strategies—especially for low-resource and long-tail languages—so production models get high-fidelity inputs.

In my work, I manage data extraction, data cleaning, deduplication, and quality measurement pipelines using Excel, SQL, and Python. I diagnose dataset anomalies, identify biases, and execute precise technical guidelines through structured feedback loops—reviewing model outputs against granular standards, running linguistic quality reviews, and generating structured feedback that drives iterative performance gains. I’m also comfortable delivering data curation simulations and diagnostic filtering assessments in remote, milestone-driven environments where reliability and rapid tool adoption matter.