I love making sense of data, and I love building software to do so. As a mechanical engineer I quickly grew a reputation as "the database guy" because I wasn't afraid to learn new tools, build reports, and write scripts to automate them. This led me down a path of web development, machine learning, and open source contribution, while learning several languages along the way.
Flask
SQL
SQLAlchemy
React
Bootstrap
jQuery
Keras
Numpy
Pandas
AWS Lambda
Bokeh
Matplotlib
Arduino
SharePoint
SAP
Microsoft Access
Microsoft Excel
Git
GitHub
Docker
A machine learning web app to track our daughter's sleep routine.
This is a complete machine learning project from data collection to production. I first connected to the API for our Google Nest camera
and collected thousands of training images. I manually categorizeed the images as "baby in crib" or "baby not in crib".
I then constructed a neural network and trained it with this data. This was a long, interative process as
I recognized various weaknesses in the neural networks (e.g. shortcuts by recognizing lighting patterns), and would then collect additional data and modify training methods to improve the model.
The project was then bundled into a web app to provide:
Browser game for popping bubble wrap. Try it out.
This is a simple game built with React to pop bubbles as fast as you can. The timer starts as soon as you start popping, and the app keeps track of your best time.
Bot for gathering COVID data and tweeting daily graphs.
Early in the pandemic I grew frustrated with the lack of trended data available for Cowlitz County.
So I wrote a script to scrape their website every day,
normalize the data to account for reporting gaps, compute 7-day rolling averages,
and tweet an updated graph of cases.
Official reporting is now improved, so this bot is no longer active.
Picture frame for playing a voicemail from a loved one.
My wife lost her grandmother unexpectedly, but was ecstatic to discover a voicemail from the previous year of her singing Happy Birthday.
To help that memory last, I integrated a touch sensor, speaker, Arduino Nano, and SD card reader into a picture frame.
When the photo is touched, the frame plays the voicemail.
API to get information for your IP address. See more.
This is a lightweight cloud-deployed API for capturing your public IP address and looking up the listed location. The API can serve a standalone HTML page for human consumption, or a JSON response for applications.
Interactive Data Visualization in the browser, from Python
I am a member of the dev group for Bokeh. I contribute general bug fixes, add tests, triage issues, and help support the community. See my code contributions here.
In my free time I work on an integration of D3-force into Bokeh to provide force-directed layout of data.
Demo: