Andrew H. Yi

Work Experience & Projects

CivicPlus (https://www.civicplus.com)

Lead Software Engineer / Senior Software Engineer

  • Ruby
  • Ruby on Rails
  • JavaScript
  • TypeScript
  • Ember
  • GraphQL

Developed municipal geospatial issue management software on the web team.

  • Developed company’s core SaaS products with a strong emphasis on Agile Methodologies, Test Driven Development, Pair Programming, and sustainable/clean architecture

Riley & Grey (https://www.rileygrey.com)

Technical Lead / Software Engineer

  • Ruby
  • JavaScript
  • Ruby on Rails
  • AngularJS
  • ReactJS

Developed with and led a team of three developers and two designers on features involving website template design, back-end services development/maintenance (API, payments, DevOps), website hosting, CI/testing/deployment and bug support (Ruby on Rails, JavaScript, Heroku).

  • Architected, developed and successfully launched the company’s first physical product – a template-based paper stationery printing service, featuring in-browser stationery editor, proofing/shipping/e-commerce systems and back-end integration with paper printing provider (Ruby on Rails, ReactJS).
  • Developed company’s core products: website hosting, domain provisioning, website templates, and RSVP/guest management.
  • Developed internal customer support tools to automate various time-consuming support related tasks.
  • Maintained, patched and upgraded all services/systems used in the company’s technology stack (Ruby, Ubuntu, PostgreSQL, Redis, etc.).

Parcel (https://www.thisisparcel.com)

Founding Engineer

  • Ruby
  • Ruby on Rails
  • JavaScript
  • AngularJS
  • Selenium

Created an integrated suite of applications including a Chrome browser extension and Web/ iOS apps focused on shopping, featuring: product bookmarking, website scraping, price tracking and budgeting (Ruby on Rails, AngularJS, Selenium, Web Scraping). .

  • Chrome Extension: https://chrome.google.com/webstore/detail/parcel-thisisparcelcom/pghnioflabfoipnhgjncpjeaogckoijf

Friends & Family Foundry (https://www.friendsfamily.co)

Software Engineer

  • Ruby
  • Ruby on Rails
  • JavaScript
  • AngularJS
  • ReactJS
  • ReactNative

Friends & Family is a socially conscious NYC based digital foundry that helps both large and small brands build and test new digital ideas extremely quickly, enabling companies to stay ahead of a rapidly changing digital world while empowering them toward a more innovative culture.

  • Buildbook: Lead team of three developers, architecting and developing RESTful JSON API (Ruby on Rails) and iOS application (ReactNative) featuring: payments, photo uploading, message/media caching, push notifications, fulltext search and project management application endpoints. Developed using TDD and Agile methodologies (RSpec, Swagger)
  • Starbucks Store: Developed various front-end widgets using ReactJS.
  • Ivy: Developed and implemented various front-end/back-end features.

Lunar Lander (OpenAI Gym) (https://gitlab.com/AndrewHYi/rl-lunarlander-showcase)

CS7642 Reinforcement Learning

  • Python
  • TensorFlow
  • Keras
  • OpenAI Gym

(Prospective Employers: This is the truncated, redacted version of the project. Please contact me at andrew.yi50@gmail.com for full project report and code.)

Trained and analyzed agents using the Double Q-learning algorithm and soft target update rule (Python, Tensorflow/Keras, OpenAI Gym)

Sleep Stage Classification (https://gitlab.com/AndrewHYi/bd4h-sleepstage-showcase)

CS6250 Big Data for Health Informatics

  • Python
  • Tensorflow
  • Keras
  • Numpy
  • Scikit-learn
  • Hadoop

(Prospective Employers: This is the truncated, redacted version of the project. Please contact me at andrew.yi50@gmail.com for full project report and code.)

Implemented and trained CNNs for single-channel EEG sleep stage classification using the National Sleep Research Institute’s Sleep Heart Health Study datasets.

Stereo Correspondence (https://gitlab.com/AndrewHYi/cv-stereo-showcase)

CS6476 Computer Vision

  • Python
  • Numpy

(Prospective Employers: This is the truncated, redacted version of the project. Please contact me at andrew.yi50@gmail.com for full project report and code.)

Implemented and analyzed several window-based stereo correspondence techniques including the sums of squared differences measure and loopy belief propagation (Markov random fields) using the Middlebury Stereo Datasets.