Amazon EC2/Lambda


Keyo Search:

Keyo is a real estate startup based out of Palo Alto, CA and Brooklyn, NY. Keyo aims to simplify the rental process for prospective renters in Brooklyn by allowing you to handle all the paperwork, interactions with the property owner and even payments.

I joined the company in Feb 2017, when the company had just formed. I replaced a full team of engineers based out of Philadelphia and took over an existing code base written by them. I built a database of over 80 million records for properties in the US and more specifically, the Greater New York City area. The data for New York gave you granular insights into the building’s history, architecture, ownership, property tax value. Generally, a search will yield the neighborhood’s demographics, quality of life and other metrics for the census bloc.

The code base is primarily written in JavaScript, using Jade/PUG to do server-side templating and rendering for every page.The database used was Postgres with the PostGIS extension, allowing one to store shape data and search through them. Additional functions were written in SQL to allow for easier search queries from the server.

The API and data search methods were developed with GraphQL, which makes searching for data on an input address analogous to traversing a graph. Address searching in the New York Area was done using PostGIS. The query sent to PostGIS would translate lat-long co-ordinates into a building parcel, which allowed for in depth fetching for any address.