Isaac Lee

I designed and developed the profile overview microservice for this open source restaurant reservation app, as well as a test suite with just over 75% code coverage using Jest and Enzyme. Using React, Webpack, Babel, and CSS modules, I prevented any possibility of CSS name collisions with other micro-services on the page, despite using Bootstrap for efficient component development and grid-tempting. Additionally, I deployed this service and a proxy server which pulled in the other restaurant… Show more

I designed and developed the profile overview microservice for this open source restaurant reservation app, as well as a test suite with just over 75% code coverage using Jest and Enzyme. Using React, Webpack, Babel, and CSS modules, I prevented any possibility of CSS name collisions with other micro-services on the page, despite using Bootstrap for efficient component development and grid-tempting. Additionally, I deployed this service and a proxy server which pulled in the other restaurant profile page micro-services using Amazon Elastic Bean Stalk. Show less

I was responsible for the design, development, and deployment of a new backend for a threaded comments micro-service on the project profile page of an open-source crowdfunding app. I compared the legacy database, MongoDB, to alternative databases by seeding 10 million fake projects into each database, and testing query times. I determined that PostgreSQL was the best database for the job in terms of query times, with the added benefit of its lightweight footprint on the legacy codebase by… Show more

I was responsible for the design, development, and deployment of a new backend for a threaded comments micro-service on the project profile page of an open-source crowdfunding app. I compared the legacy database, MongoDB, to alternative databases by seeding 10 million fake projects into each database, and testing query times. I determined that PostgreSQL was the best database for the job in terms of query times, with the added benefit of its lightweight footprint on the legacy codebase by maintaining a document-oriented schema using PostgreSQL's JSONB datatype. This came with the added benefit of much higher read-write consistency than the legacy database. I then deployed this new backend to a free tier AWS server and simulated web-scale traffic. To minimize request times, I implemented an NGINX loadbalancer across two free-tier AWS servers to demonstrate that horizontal scaling could reduce page-load times from 2000ms to 200ms, under the load of 1,000 requests per second. Show less