Experience

Pave

• United States
• Software Development
• 100 - 200 Employee

• Data Scientist

  • Oct 2021 - Feb 2023

  San Francisco • Created detailed understanding of baseline product health and then collaborated with stakeholders on decisions such as feature release timelines by designing a practical, comprehensive guide rail metric based on customer usage statistics, expert subject knowledge, and simple heuristics • Proactively identified quality issues with the raw data, the data architecture, and the machine learning models before updates to users by building continuous monitoring of product health over time… Show more • Created detailed understanding of baseline product health and then collaborated with stakeholders on decisions such as feature release timelines by designing a practical, comprehensive guide rail metric based on customer usage statistics, expert subject knowledge, and simple heuristics • Proactively identified quality issues with the raw data, the data architecture, and the machine learning models before updates to users by building continuous monitoring of product health over time with a thoughtfully architected process in dbt, robust to product changes, that transformed data daily from first BigQuery input to summarized output for internal user viewing • Abruptly decreased relevant negative outcomes for users, raising the product health metric on the order of a percent, with a holistic evaluation of model training data that included recommendations for process revisions intended to scale • Acted as lead IC for the overall merging of Pave data with acquired Option Impact salary data; met company deadlines and OKRs while keeping minimal data loss and constructed a reliable data pipeline with Apache Airflow and GCP Composer • Oversaw data team work to reach feature parity with Option Impact, including reaching agreements with product and frontend software teams, assigning data-related tasks/JIRA tickets, mentoring teammates, leading scrum/standup project updates, and making external vendor purchasing decisions Show less



Fivestars by SumUp

• United States
• Financial Services
• 100 - 200 Employee

• Product Data Scientist

  • Apr 2020 - Aug 2021

  San Francisco, California, United States • Spearheaded a customer segmentation analysis resulting in a more accurate understanding of total addressable market, then evangelized the results through presentations to cross-functional stakeholders and consulted with varied departments on practical usage, creating company-wide adoption • Prevented the waste of hundreds of consumer-hours by independently considering the consumer funnel, identifying a major pain point from uncollated product alpha data, and clearly communicating the… Show more • Spearheaded a customer segmentation analysis resulting in a more accurate understanding of total addressable market, then evangelized the results through presentations to cross-functional stakeholders and consulted with varied departments on practical usage, creating company-wide adoption • Prevented the waste of hundreds of consumer-hours by independently considering the consumer funnel, identifying a major pain point from uncollated product alpha data, and clearly communicating the findings to product owners, leading to design changes and isolation of a third party flaw in new hardware • Led quantitative measurement for merchant facing payments products, using statistical methods such as A/B testing, survival analysis, and heuristics-based models, contributing to an organizational 5x increase in revenue • Immediately improved the speed of ad hoc requests and facilitated company data literacy by modernizing the ETL process serving as the basis for over 50% of company dashboards, achieved through writing efficient Presto queries to transform billions of rows from a data lake into optimized Redshift tables used for Tableau • Preemptively ensured that OKRs were met across multiple teams by strategically advising product manager and engineering lead partners with actionable recommendations, made possible through driving analytics instrumentation corrections and monitoring KPIs for anomalous behavior • Conducted deep dives on the health of the integrated software products by synthesizing knowledge from other subject matter experts, querying data to understand both the consumer and merchant perspective, and surfacing relevant opportunities, facilitating increased prioritization of the project Show less



Insight Data Science

• United States
• Higher Education
• 1 - 100 Employee

• Insight Data Science Fellow

  • Sep 2019 - Apr 2020

  San Francisco • Cleaned and collated tens of GBs of compressed JSON files representing real-world data using AWS, mining ~4 million relevant rows, as a precursor to feature development and ML model construction • Developed a binary Logistic Regression model from data containing highly imbalanced classes, ultimately achieving a recall of 89% in labeling a class representing less than 1% of total data • Built The Banalyzer, a Flask and Python-based web app providing social media users of Reddit… Show more • Cleaned and collated tens of GBs of compressed JSON files representing real-world data using AWS, mining ~4 million relevant rows, as a precursor to feature development and ML model construction • Developed a binary Logistic Regression model from data containing highly imbalanced classes, ultimately achieving a recall of 89% in labeling a class representing less than 1% of total data • Built The Banalyzer, a Flask and Python-based web app providing social media users of Reddit with an estimate of subreddit sentiment by predicting the likelihood of subreddit banning Show less



Purdue University

• United States
• Higher Education
• 700 & Above Employee

• Graduate Research/Teaching Assistant

  • Jul 2012 - May 2019

  West Lafayette, IN • Utilized Perl and R to perform hypothesis tests on hundreds of astrophysical phenomena exhibiting stochastic behavior over unequally sampled time intervals, providing evidence for certain magnetohydrodynamics-based theories • Analyzed experimental data within a MySQL database, found inconsistencies to be corrected, and determined which qualities and trends in the data were indicative of observational bias due to differing orientation, distance, and energy distributions among… Show more • Utilized Perl and R to perform hypothesis tests on hundreds of astrophysical phenomena exhibiting stochastic behavior over unequally sampled time intervals, providing evidence for certain magnetohydrodynamics-based theories • Analyzed experimental data within a MySQL database, found inconsistencies to be corrected, and determined which qualities and trends in the data were indicative of observational bias due to differing orientation, distance, and energy distributions among astrophysical jets • Developed weekly, 60 minutes presentations in order to communicate difficult physics concepts in simple terms to entry-level students

• Women In Science Programs Leadership Team Member

  • Jul 2016 - May 2017

  West Lafayette, IN • Collaborated with a group of multidisciplinary scientists, as department representative on the Women in Science Programs Leadership Team, to manage monthly professional events consisting of speakers sharing their professional expertise with students



Murray State University

• United States
• Higher Education
• 700 & Above Employee

• Undergraduate Researcher

  • Aug 2011 - May 2012

  Murray, KY • Coordinated a series of computationally difficult processing jobs designed to find significant time series patterns within a large data set • Successfully composed a grant application leading to new research equipment and travel accommodations

• Teaching Assistant

  • Aug 2011 - Dec 2011

  Murray, KY • Guided students through software relating astronomy on a grand scale to knowledge of their local sky



Purdue University

• United States
• Higher Education
• 700 & Above Employee

• Undergraduate Researcher

  • May 2011 - Aug 2011

  West Lafayette, IN • As part of a summer research experience, adapted instrumentation parameters on a ray-tracing modeling program in order to improve simulation of a future telescope



Murray State University

• United States
• Higher Education
• 700 & Above Employee

• Tutor

  • Aug 2008 - May 2011

  Murray, KY • Taught student-athletes of varying skill levels, individually and in small groups, ensuring they met GPA goals on general subjects