Architecting and optimizing real-time, event-driven AI data pipelines in Go, enhancing observability and explainability for large-scale unstructured data processing. Building high-throughput systems with gRPC, Kafka, and ClickHouse, enabling scalable infrastructure for ML model deployment and monitoring. Driving technical strategy with the founding team, shaping product direction and ensuring reliability, scalability, and performance from prototype to production.

Implemented advanced ML models (PyTorch, TensorFlow) for high-dimensional data analysis in food tech R&D, improving predictive accuracy of product performance by 25%. Deployed scalable APIs (FastAPI, gRPC) and CICD pipelines on AWS (Terraform, Docker), reducing model iteration cycles from days to hours. Led cross-functional collaborations, integrating complex ML pipelines with business metrics, contributing to a 15% reduction in production costs.

Developed Python and PyQGIS scripts automating wind turbine layouts, enabling data-driven site planning and boosting renewable energy output efficiency by 30%. Implemented large-scale data ingestion and transformation pipelines (Apache Spark, Airflow) to handle multimodal datasets, accelerating environmental simulations by 40%.

Pioneered research on non-independent component analysis and interpretability in algebraic statistics for complex ML systems. Published findings in top-tier statistics journals (e.g., under review at Algebraic Statistics), presented at international conferences.

My research focused on algebraic machine learning and statistical theory, particularly on non-independent component analysis and graphical models. This work led to the publication of Partitioned Independent Component Analysis, which is available on arXiv.

We prove that the Gorensteinification of the face ring of a cycle is totally p-anisotropic in characteristic p. In other words, given an appropriate Artinian reduction, it contains no nonzero p-isotropic elements. Moreover, we prove that the linear system of parameters can be chosen corresponding to a geometric realization with points on the moment curve. In particular, this implies that the parameters do not have to be chosen very generically.

Our research focused on the differential power operation on ideals. We identified a class of monomial ideals in characteristic 0 whose differential powers are eventually principal. We also explored the containment problem between ordinary and differential powers of ideals and introduced a novel closure operation, called differential closure, which agrees with taking the radical of an ideal in simple D-modules.

I collaborated with Art Duval from UTEP on applying the Abelian Sandpile Model to DDoS mitigation. This project led to an ongoing publication titled, 'Analyzing Network Topology for DDoS Mitigation Using the Abelian Sandpile Model.'

I worked in the Experimental High-Energy Particle department, contributing to the LDMX project. My work included SketchUp design, data collection from oscilloscopes, scintillators, and PMTs, and data analysis using Python. The project also involved designing and modifying electrical circuits.