San Diego, California, United States
• Executed ETL operations on 5000+ rows robot data from GCP BigQuery using Python
• Applied ML dimension-reduction techniques like PCA, UMAP, and t-SNE for visualizing error clusters and diagnosing robotic autonomy performance abnormalities
• Optimized SQL queries to isolate the abnormal robots routes influencing data distribution and constructed efficient Python APIs, resulting in a 10% improvement in web app load time
• Implemented Airflow and Spark workflows to generate daily error shift images and distributed them to cross-functional teams for analysis
• Collaborated with a 10-member team, utilizing Git and Jira for streamlined task management