At DataRobot, I rated AI and LLM outputs for accuracy, factual alignment, tone, and policy compliance. I also labeled datasets and ranked model responses to support RLHF pipelines.
I improved data quality metrics and task efficiency by 35% through Python scripting and workflow optimization. I also refined evaluation rubrics and annotation guidelines.
At Scale AI, I processed text and audio datasets for NLP and speech recognition applications. I generated prompts for language models and wrote Python scripts to automate data cleaning and quality checks.
At Stanford University, I developed machine learning models for classification and pattern recognition, and analyzed research data using Python and R. Earlier, as a Technical Assistant at LXT AI, I prepared and annotated datasets for speech-to-text and text-to-speech systems.

