Continuing to support several product teams that focus on user safety via data analysisLed a strike team of 5 data scientists to investigate a significant and sudden baseline shift in one of our key metrics to identify which products and policy changes were responsible, and which should obtain more resources for future improvementsCreated a data catalog for common metrics and data sources used by our science team, which expedited several future audit requests from our Data Infrastructure teamLed an investigation into a new attack vector, and identified several new friction points our product teams could take advantage of to reduce on-site harm, while not affecting legitimate users

Supported several product teams that focus on user safety via data analysisLed an analysis comparing the performance of two potential vendors that could support one of our most frequently used moderation productsHelped develop several heuristics to identify bad faith users for our ML teamsImproved communication and collaboration between two data science teams to speed up response time for incoming site incident investigation requests

Provided data quality support for the Central Analytics teamMaintained monitoring scripts for Twitch’s most frequently used eventsCreated incident reports to summarize data quality incidents, and their impact to Data ConsumersModerated communication between Data Producers their Data Consumers to talk about known incidents, and share how data is being used in different parts of the businessLead an effort to test adoption of new data cataloging software to make it easier for SMEs to share their knowledge

Provide data support for the Anti-Money Laundering Data Management teamMaintain and improve a data pipeline to identify and flag fraudulent entities applying for a small business credit card Developed a data validation method for the Amazon Web Services Case-Study team using Python, Teradata, and Redshift SQLCreate and maintain cloud migration strategy documentation detailing the difference between cloud and on-prem solutionsManage several Teradata data containers used by over 100 analystsWork with analysts in pair programming exercises to improve team knowledge of Python and AWS

Provided data support for Consumer Rewards and Swap My Card programsAssisted in the launch of the Quicksilver product with user testing and monitoringIdentified negative customer experiences and recommended paths of account remediation to business team

Worked on a multidisciplinary team of engineering students to develop an accurate modeling tool for our sponsoring companyModeled process using Aspen Plus, Visual Basic, and Fortran

Contacted patients regarding appointment informationOrganized incoming and outgoing medical information for patients