Geoff Rollins

Data Scientist | Data Engineer | Biophysics PhD

Professional Summary

  • Data Scientist with strong quantitative background that spans data science, data engineering, and physics
  • 4+ years experience working in a fast-paced tech startup environment with extensive remote experience
  • 10+ yrs experience with Python; 4+ yrs experience with R and SQL; 2 yrs experience with web frameworks (django, angular)
  • Passionate about refactoring legacy code to mitigate technical debt, guided by tests
  • Effective communicator – experienced in writing and presenting reports to both business and technical audiences

Work History

Data Scientist | Atlassian

2020 - present
  • Trello data science team

Software Engineer in Test | DataStax

2018 - 2019
  • Tested Apache Cassandra at scale using Python, Jenkins and in-house tools
  • Built custom dashboards and data pipelines in Python and R to make it easy for PM and dev teams to monitor test results and catch regressions early
  • Safely refactored and added new functionality to a legacy ETL tool for extracting data from Jenkins and loading results into test tracking platform

Data Architect | Dropbox

2017 - 2018
  • Refactored and optimized key data pipelines in Hive and Luigi to stabilize and speed up operational reporting for Customer Analytics team, eliminating significant technical debt along the way
  • Built dashboards using in-house Tableau-like platform to track Dropbox Help Center traffic

Data Scientist | Dropbox

2014 - 2017
  • Led the development of a new customer satisfaction survey that influenced overall company strategy. My work covered all phases of the project: questionnaire design, population sampling, statistical analysis, and presentation of findings and recommendations to diverse set of stakeholders
  • Built predictive models of customer churn and worked with a cross-functional team to A/B test marketing campaigns to reduce churn; also built reporting pipelines and dashboards to track churn

Scientific Software Developer | Freelance

2013 - 2014
  • Developed web apps for academic clients to communicate scientific findings to a wider audience using Django and AngularJS deployed on Google Cloud and Digital Ocean.
  • Developed Sophia, a GUI-based molecular simulation tool written in Python (see below)

Work Samples


Open Source

Sophia is an educational molecular simulation package. The goal of Sophia is to provide a gentle on-ramp for newcomers to all-atom molecular simulation methods (energy minimization, molecular dynamics, simulated annealing, monte carlo sampling, etc.). Production simulation tools, like Gromacs and OpenMM, are designed for performance, whereas Sophia is designed as an interactive teaching and learning tool for small-scale simulations. Sophia is implemented in Python (~5K lines) as a plugin to UCSF's Chimera.