Evelyn Marcel

I am an analytical software and data professional focused on AI model training and evaluation, with hands-on experience assessing language models for accuracy, coherence, and reasoning quality.

I evaluate and rate AI-generated responses across domains including science, math, finance, and customer support, annotate datasets for supervised fine-tuning and RLHF, and participate in multi-turn conversation tasks that simulate real-world interactions.

I combine technical skills (Python, SQL, Git, HTML/CSS, React, FastAPI) with strong linguistic proficiency and critical reasoning to identify logic gaps, tone inconsistencies, and factual errors in model outputs.

I have completed large-scale annotation projects with high QA ratings, developed a personal annotation quality rubric to improve speed and accuracy, and deliver structured feedback that improves model alignment and user experience.