At Readcoor, I've worked on instrument control software for an in-situ sequencing platform. My responsibilities ranged from high level logic in Python3, embedded firmware running on custom PCBs, as well as user interface and design work done in React.

At MIT, I worked with hundreds of terabytes of video and tabular data distributed across multiple remote servers and databases to assist autonomous vehicle research. Data-driven scripts that I've developed in Python include backups, cleaning and preprocessing pipelines, computer vision detection algorithms and GPS visualizations of thousands of unique driving trips. I have presented my work in driver head pose classification, instrument cluster icon detection, crash profiling vehicle kinematics and cognitive load detection to sponsors on four separate occasions, helping to secure hundreds of thousands of dollars in sponsorship for the lab and contributing to cutting edge research. Technologies that I've used here include Python, PostgreSQL, C/C++, Bash, Linux (Ubuntu & CentOS), LaTeX and Markdown.

Classified 4 masonry worker tasks from accelerometer data to increase construction worker safety.

Developed KNN, SVM and Naïve-Bayes classifiers in MATLAB.

Extracted features using statistical analyses on signal waveforms.

Plotted data on visualization tools to help visualize separation of data into action classes.

Helped design and set up data streams for a program that allows you to type with your brain.

Neurological and eyetracking data were streamed and analyzed on local network.

Projects involved a combination of MATLAB, C++, C and python programs.

EEGLAB was used to help analyze EEG data using provided machine learning methods.