Create image processing pipelines for efficiently ingesting, analyzing, and performing ML inference on large whole-slide image datasets.Implement numerous computer vision deep learning networks for generating predictions and segmentations from whole-slide images. Processed and refined whole-slide image features using classic and state-of-the-art deep-learning computer vision approaches. Measure and optimize deep network training across multiple GPUs on local machines and in the cloud.

- Awarded a National Science Foundation Graduate Research Fellowship totaling US$141,000- Established an image processing pipeline for a multi-class semantic segmentation of thousands of images, combining results from a random forest pixel classification network and a custom deep U-Net network architecture to produce high-resolution segmentation of numerous tissue types. - Created a data processing and presentation pipeline using TFRecord parsing to improve network training speed from several days to less than an hour.- Designed a machine learning algorithm to aid in semi-automated identification of arrhythmias within ECG datasets. - Combined histological tissue processing and computer vision to create a pipeline for generating sub-µm, data-rich 3D reconstructions of important cardiac tissue structures.- Created a semi-automated data acquisition protocol for gathering light-scattering spectroscopy data from tissue surfaces using a custom probe, increasing sample sizes from ~200 to more than 20,000.- Applied statistical methods to determine relations between complex spectral datasets, similar in structure to omics datasets.- Regularly communicated complex research findings and methods with diverse stakeholders, including surgeons, engineers, professors, students, and industry partners. - Coordinated the improvement of tissue sample embedding, sectioning, staining, and imaging protocols, improving percent of usable tissue samples and images from 30% to more than 90%. - Developed two separate neural networks capable of predicting tissue characteristics from light-scattering spectroscopy data gathered from tissue surfaces. Published in Journal of Biomedical Optics (JBO) and Sensors.- Refined and adjusted protocol for identifying and segmenting ovine cardiac conduction system tissue in histologically stained tissue samples. Published in PLOS One.

- Taught numerous labs on building and using ECGs and other physiological sensing equipment, modeling of electrical activity in the heart, and acquiring, processing and interpreting physiological data. - Mentored more than two dozen graduate students in experimental design, programming practices and techniques, digital signals analysis, and data acquisition.

- Created a novel, decentralized, redundant manufacturing data and failure logging system using edge computing systems, SyncThing networking protocol, and Linux devices. - Designed, built, and maintained fixtures used in a high-throughput production line to resolve manufacturing issues, improve ergonomics for manufacturing staff, and inspect produced devices for defects. - Performed investigations into device failure mechanisms, including gauge R&Rs, and recommended alterations to manufacturing protocols to reduce failure rates.- Coordinated implementation and maintenance of equipment documentation and calibrations.- Refined and improved R&D device production protocols for successful implementation in a mass production setting.- Maintained code for on-site custom built production machines and electronic fixtures.

- Created a novel pressure-sensing insole worn in normal shoes/sneakers as an experimental mobile gait lab. Refined the circuit design, manufacturing process, and data analysis. Circuits were designed in Eagle. - Built software (with accompanying graphical interface) enabling visualization of time-series data by time-aligning data with independently acquired video, and providing several user-selected methods of plotting or visualizing the data. Written in python and MATLAB.- Designed a novel, sterilizable medical device used to measure pelvic floor strength in women. Used daily in a large study on postpartum pelvic floor disorders, patent currently under review. Designed in SolidWorks. - Developed and refined a quality assurance protocol for detecting mechanical failures in catheters used to treat lung disease. Presented results at the senior undergraduate research symposium and was awarded Best Presentation.- Created lab website and performed graphic and web design.- Designed and built a collapsible device used to measure abdominal strength, used in a study on pelvic floor disorders. Device design was featured in the university entrepreneurship periodical.- Wrote standard operating procedures for manufacturing pressure-sensing insoles.- Manufactured experimental catheters for use in the diagnosis and treatment of lung diseases. - Carried out general electronics and rapid prototyping and including wire bonding, soldering, 3D printing, over-molding, and general CAD.

- Created a graphical visualization of geographic customer data demonstrating trends in location and quantity.- Demonstrated advanced soldering and electronics manufacturing skills.- Gathered and analyzed data on device manufacturing and electronics failures.- Created vector graphic images of products in adobe illustrator.

- Developed a novel method of organization and accountability for group video editing. - Refined dozens of MBA courses to ensure maximum benefit and retention for the students.- Edited hundreds of hours of video footage to ensure a professional and seamless presentation of lecture materials. - Participated in creating training material for future employees. - Created animations and other supplemental material to enhance the value and effectiveness of courses.

- Coordinated directly with clients about personalized logo and website designs.- Designed and animated commercials and graphics for local and national companies.- Content creation and photo editing to enhance quality of client websites and social media presence.

- Coordinated all event teams and catering. - Orchestrated setup and takedown and of events. - Interacted with clients staff as the first and primary point of contact.