Jacob Bieker
Astrophysicist & Photographer
I am currently a Senior Machine Learning Engineer at Vida, where I focus on geospatial foundation models, climate risk for infrastructure, and climate forecasting models. I was previously a Senior Machine Learning Research Engineer at Open Climate Fix, where I focused on developing and deploying models for predicting satellite data and solar energy output. I previously was at Scale AI, where I focused on building models for 3D use cases, primarily with LiDAR and image data. I graduated with my BSc in Physics and Computer Science from the University of Oregon, and my MSc in Astronomy and Data Science from Leiden Observatory.
I use computers to fix climate change, currently through creating a better understanding of climate risk for infrastructure, as well as on bettering weather forecasts through ML. I worked to make better predictions for solar energy nowcasting so the elecrical grid can be more efficient. I research galaxies, especially high redshift ones, to discover how our Universe came to be. I also apply machine learning to astronomy to explore its applications. In my free time, I hike, ski, and shoot landscape and night sky photography around the world.
I use my interest in computing and interdisciplinary research to study massive sets of data and produce insights in various contexts, including astronomy, biology, and Earth observations. I've done this at Google, NASA, Leiden Observatory, TU Dortmund, Scale AI, Open Climate Fix, Martian, and now Vida.
My research interests are focused on interdisciplinary research, especially in astronomy, but also biology and computer science. I've done research at NASA on hyperspectral data and productionized machine learning algorithms for Google's Gsuite. At Scale AI, I built up models for prelabelling autonomous vehicle data and linting the outputs from human taskers to improve the human-in-the-loop system. Some of my current and past research is shown below. My CV is available here, ADS here, and Google Scholar
My photography is primarily based around landscapes, the night sky and pushing the technical limits of the camera and computing while creating gigapixel images.