I've shipped 90+ production computer-use tasks for Fleet AI across tools including Gusto, QuickBooks, DocuSign, Jira, HubSpot, and Looker. I write prompts and pass/fail checks from JSON snapshots, turning product-review feedback into shipped improvements quickly.
At Alignerr and Outlier, I evaluated multi-turn Python coding trajectories and LLM responses against detailed rubrics. I also reviewed customer-service dialogues, rewrote SOPs, and improved feedback templates for consistent quality.
I built a COVID-19 survey analysis pipeline and Streamlit dashboard using Python, Jupyter, CSV/Parquet, Docker, and bilingual documentation. My operations background includes Excel inventory systems, receiving logs, document control, and last-mile dispatch at Walmart de México y Centroamérica.

