Phobos• A primary developer for the Phobos diagnostic program, a subproject of the MARS scheduling system (latter now in operational use by the U.S. Space Force). Contributed significant design decisions (what to monitor, how to monitor, how to integrate components, and how to display information) and core implementation (see below)• Implemented monitoring for machine resource usage, a program GUI for filtering and displaying diagnostic results, private network communication both for Phobos-Phobos and Phobos-MARS messaging, and a variety of smaller components, integrations, refactorings, and bugfixes in Java• Collaborated with a development team by participating in daily stand-ups, coordinating tasks with Jira tickets, and performing regular Pull Request (PR) reviews on BitBucketInFormer• Key developer for InFormer, a project for the Dept. of Energy intended to forecast human-induced earthquakes• Created data pipeline with Numpy, Pandas, and PyTorch that aggregated data from multiple earthquake anddrilling datasets, applied engineered feature functions (definable by other developers via a query tool I also created), and fed feature data into a Graph Neural Network (GNN) for training.• Collaborated with team by eliciting design requirements from team leads, communicating with consulting Stanford Geophysics professors about data processing, and using GitLab to submit Pull Requests (PRs)