I solve busines problems through data science and software engineering. Beginning with the MVP, I iterate using profiling and testing to achieve performant systems and derive actionable insights based on well-tested and well-documented code. Experienced solving complex geospatial data challenges through cross-team collaboration, I can draw on ML and statistics in my engineering. With two PhDs in math and music I take an agile approach to solving novel challenges that increase business value.
Personal projects:
- Actively building Cap'n Proto Exporters and Receivers in Rust that integrate into the opentelemetry ecosystem. See opentelemetry-otlp-capnp.
- Co-authored the Rust crate
iterative_methods- (Weighted) Reservoir Sampling implemented, tested, and demonstrated with visualizations.
- YAML adaptor for writing data to file as part of an iterative method
- Unscrolled
- A prototype UI for AI assistants with collapsible responses that makes it easier to navigate long exchanges.
- Most of the code was generated using Claude Desktop to rapidly develop a demo. This is not intended for use but only to demo the kinds of features that I think will make AI Assistants more useful.
- Gender bias in music genre labels in Wikipedia
- 2nd Place Award for NYC Digital Humanities Graduate Student Project
- χ-square test showing that female artists are significantly under-represented among all artists with >5 genre labels
- Scrape, clean and normalize culturally complex text labels (genres) from Wikipedia
- Based on a Kaggle dataset of 15K musicians with gender labels