Hello, my name is Nathan
I am an IO Psychologist specializing in Assessment Development and End-to-End Machine Learning.

I used machine learning to develop 75 algorithmic alternatives to key products predicting employee performance.

I led a team examining the criterion, construct, and content validity; reliability; and, adverse impact/group differences of the final algorithms. I modified models to reduce bias as needed based on domain expertise.

I streamlined a focus group note coding process using natural language processing.

I used a group of subject matter experts (SMEs) to establish a human benchmark for comparison. Final models identified personality relevance more accurately than SMEs, and reduced coding process time by a factor of 90.

I developed a full-stack facial recognition application using the API from Clarifai. Users register for the application and are then able to upload images. The Clarifai API identifies the location of any faces in the image, and the application draws boundary boxes around them.

The front-end is built using React.js and includes both sign in and registration pages. The back-end was built using Node.js and Express.js with a PostgreSQL relational database to keep track of user names, emails, and hashed versions of their passwords.

I built a secure password checker in Python using hash functions.

The script converts the password to be checked to SHA1 hash. The first five characters of the hash password are sent to the Pwned Passwords API. The API checks all hashed passwords with the same first five characters. API matches are sent back to the script which identifies whether the password was previously leaked by matching API results to the entire SHA1 hash stored locally. Pwned Passwords API never sees the password, completely secure.