Michael R. Email & Phone Number

- Lead author on four published machine learning patents relating to real-time anomaly detection.- Responsible for research, development, and deployment of patented machine learning systems forflagship product.- Led software and machine learning development for projects delivered to Fortune 100, US federal,and international clients.- Built data processing pipelines to efficiently manage millions of data points and generate statisticallyoptimal data sets for training.- Completed machine learning research that won all three phases of competitive funding from the USDefense Advanced Research Projects Agency (DARPA).- Developed real-time machine learning infrastructure for NVIDIA embedded computing boards,cloud, and on-premise deployments.

Bellingham, Washington, Us

- Designed and implemented custom deep and recurrent neural network architectures to model sports data.- Created recurrent neural networks that outperformed baselines at forecasting timeseries data (base- ball plays).- Created a custom LSTM cell state initialization solution to better model individual sports teams. Built a data-processing pipeline that interfaces with a database and normalizes or embeds features appropriately.- Created custom models with convolutional, LSTM, and deep neural network architectures. - Visualized and presented statistical discoveries.

Bellingham, Washington, Us

• Developed a program in Java to automate a common microscope imaging task. This reduced task time from around forty minutes to near instantaneous.

• Developed a system for recording and interpreting data visually and statistically from visitor surveys.• Authored an eighty-page booklet detailing the scientific concepts behind each exhibit.• Created and updated graphic design for public documents.