- Developed high-performance satellite constellation mission-planning software that reduced planning runtime by over 90%- Designed MLOps architecture and trained computer vision models for feature detection in satellite imagery covering hundreds of thousands of square-kilometers

Developed marketplace for Credit Unions to buy and sell securities using Kafka-backed event-drivenarchitecture and DynamoDB single-table design

Responsible for leading architectural direction of the BlackSky Spectra AI Platform, developed by a team of 10+ backend engineers.Led overhaul of public API for ordering imagery and analytics, reducing friction and number of requests made on Customer Success team.

Designed and Implemented AWS-based pipeline architecture for satellite imagery analytics, from customer-order through computer vision model inference and product delivery.Lead backend-developer of new application to support customers in managing, ordering, and scheduling of large satellite-constellation imagery collections.

Developed software architecture for streaming geospatial-temporal data, and scaled up prototype machine learning algorithms using Spark, Oozie & HBase.Developed a daily global change-detection raster product using NOAA's VIIRS satellite imagery to automatically detect and visualize the extent of power outages around the world.

Perform and lead data analysis projects for the FAA and international aviation organizations, analyzing terabytes of geospatial-temporal data.Completed and delivered software tool to the FAA that guides Wake Vortex Recategorization improvement, which saves aviation industry millions of dollars per year.Led effort to educate analysts in using Spark, and best practices for reproducible data analysis.

Developed from scratch a resource planning web application to allow program managers to effectivelyassign data scientists to projects where they fit best. Tech stack: NodeJS, AngularJS, MongoDB.

Completed predictive analytics project from data ingestion through deployment, using Agile methodology.Analyzed 75 million software usage logs using Apache Spark and scikit-learn to predict abnormalterminations.Received weekly analytics classes from world-class data scientists.

Developed a JavaFX/Javascript application that maps telemetry data. Used in a successful customer demo for a large, international client.Automated issue-tracking analysis and metric production process. Cut process time from one hour of manual-intensive work per week to one hot-key press in Excel.

Assisted professors with instructing an introductory course in Object-Oriented Programming in Java

-Marching Band Instructor-Sectional Coach-Private Teacher

Taught classes in reading, writing, speaking, and culture to students at Konkuk University in Seoul, South Korea as part of the Internship and Cultural Experience program.