About

My name is Michael, a Data Scientist who enjoys building at the intersection of AI, physics, and biochemistry in topics such as High-Field NMR, Crop Yield Detection, Polygenic Risk Scoring, and Graph Neural Networks.

At NIH PRIMED, my team encountered a significant challenge with the AnViL platform we built (https://anvilproject.org/). It lacked efficient job orchestration, hindering our ability to manage data workflows at scale, such as Bulk RNA Seq. As a biochemist in undergrad, I also noticed that my peers frequently struggled to create effective simulations for iterating on experiments such as NMR Spectroscopy due to ineffective yields or lacking sample data. While AnViL is often seen as the "Snowflake for Biotechs," it doesn't fully address the needs of users as its lacking core functionalities outside of data warehousing.

While at NASA, this workflow bottleneck prompted us to build a 'Snowflake for Agriculture.' With a large-scale actionable geospatial data warehouse for agriculture, we could then train advanced AI models for crop yield prediction using satellite imagery, reducing the impact of ineffective agricultural practices. Additionally, our SpaceX colleagues faced similar challenges in synchronizing the "wet lab" aspects (physical testing of materials and components) with the "dry lab" aspects (AI-driven simulations for design optimization and performance prediction).