Sofia Baig

I’m a freelance data annotator focused on ensuring language-model outputs are high-quality, accurate, and aligned with specific linguistic criteria. In my current work, I evaluate large language model responses for quality, correctness, and adherence to requirements such as atomicity and localization.

I also support model performance through prompt engineering—developing and refining prompts to improve response quality. To keep datasets reliable, I conduct high-precision document comparison and validation for dataset integrity.

Beyond evaluation, I systematically categorize and tag complex data to train machine learning models. I take pride in careful, detail-oriented annotation that translates language and context into consistent labels.

Before data annotation, I built strong people-facing and production skills through tour guiding and freelance photography. Giving tours to school groups and private visitors, facilitating workshops, and managing information points strengthened my communication, while portraiture and event photography—including 3 self-published online magazines—honed my creative and project-management mindset.