UI & ML Developer

Latest Work

API Dashboard

The Data Dashboard provides an API to multiple internal databases for the client. I was the sole javascript developer on their new UI that uses the API to generate charts and allows the user to download the data as a csv or excel file. The focus was to allow for new data types to be quickly added to the application, and new charts as new use cases arise. Speed and ease of use were also key.

  • React.js/Redux/Material UI
  • Handling large data sets in the browser (100+ mb)
  • Plotly.js charts
  • Swagger/Open API
  • Azure Authentication
  • AWS Cloud Formation + AWS ECS

Power Meter

Power meter took a video, and later a live stream from a webcam, of a lifters workout and measured the amount of work being done. I used react.js, tensorflow.js, and a custom webpack plugin to first build a testing platform for model accuracy and later build an app that streamed the webcam and statistics about the workout.

Steward AI

Greenhouse monitoring hardware and software that used images collected from a raspberry pi to analyze plant growth. Farmers were given a dashboard to monitor plant growth over time.

3 Farmers signed up for a private beta that ran from 2018 to 2019.

Last Stop

Firebase + Angular application for connecting delivery drivers to customers in an easy and simple workflow. Drivers get an app with a couple buttons to update the customer on their location, automatically sending an ETA. Customers get emails with updates and a link to a real time view of where the driver is.  Managers get a dashboard with all ongoing routes and alerts for routes that are behind schedule.

Image Annotation Platform

The client processes invoices for restaurants using a combination of machine learning and data entry people. The backend has a data pipeline that combines machine learning and manual data entry. I was the lead developer for the front end that allowed the data entry team to process the invoices and annotate the invoices to improve the machine learning models. It included:

  • image label selection
  • workflow management
  • industry specific business logic
  • collecting data from micro-services, pipeline API, and document APIs to validate user input