Brian Mcgill Email and Phone Number

- Oversees Business Analytics, Reporting, Machine Learning, and Data Engineering- Part of executive committee in charge of developing Direct-to-Consumer solar marketing models- Led development of a real-time data transfer product from Postgres to Looker via S3 and Snowflake- Revamped stipulations policy for solar product, resulting in a 50% reduction in stipulations with minimal credit risk and a $10M/annual increase in disbursed loans

- Led the development of a scalable analytics infrastructure leveraging Prefect, Kubernetes, DBT, GitLab, Snowflake, Looker, and AWS- Built first internal Machine Learning Models for use in student loan refinance pricing and cash flow projections- Oversaw a credit expansion in the student loan refinance product leveraging reject inference data from Transunion, resulting in a 5% increase ($30M) in annual originations- Developed 1st generation credit policy for the new solar lending product ($100M annual originations)

• Developed MSA level pricing framework to provide optimal price based on regional market competition (estimated $30MM - $40MM increase in annualized NPV)• Head model developer in charge of implementing COAF’s account-level pricing framework with estimated $10B in annual originations• Functional lead of two junior-level Data Scientists supporting Customer Office and Collections• Served as lead DS for COAF collections managing one junior-level DS with focus on the development of an intelligent contact strategy with annual operations savings of $20M

- Lead DS in the development of a model for optimal vehicle repossession timing using an AWS/Python-based machine learning algorithm (expected annual savings of up to $40MM)- Developed Capital One's first intelligent collections targeting model (expected roll out June 2018) which will be used to prioritize accounts for Capital One's Auto Finance delinquent portfolio-Head DS in charge of DOE for a digital marketing pod utilizing Google AdWords, YouTube, Contextual and Programmatic Display campaigns- Led meetings with Sr Directors in order to brainstorm and prioritize DS projects that are used by the DS and Auto VPs for project prioritization- Established an AWS based infrastructure for use in the development of an internal marketing model using 3rd party data from Neustar’s DMP (~3.8 TB/day)- Developed a predictive model using clickstream data that determines the likelihood of Auto Navigator applicants using the product. Used by the call center to contact customers that unlikely to use the product.- Project lead on three experimental design tasks aimed at improving the Auto Finance division marketing campaign- Identified and proposed a new method of selecting the correct trade line for refinancing that reduced the error rate from 23% to 2% in a $4 billion line of business

Developed an equation based pricing method to replace the currently implemented table based approach

- Taught recitations for two undergraduate introductory statistics courses aimed at liberal arts majors and two recitations for engineering and math majors- Indepently taught the introductory course for engineers and math majors (designed syllabus, lecture notes and exams)