Opendoor Labs

Engineering and Data Science Blog
Latest articles
  • The Two Cultures of Machine Learning Systems

    A tale of woe

    Read more
  • Using Postgres window functions to link data across many sources

    Let’s say you want to know about properties in Phoenix. The first thing you would do is look around for some data about these properties, and you would indeed find a lot. You would find realtor listing databases, tax assessor records, recorder office data as well as data from an endless array of third-party providers, and pretty quickly, you would find yourself inundated with heaps and heaps of data.

    Read more
  • AMI rolling update using Ansible

    This post covers how to achieve zero downtime updates of an AMI with an AWS Auto Scaling Group and using Ansible. At Opendoor, we use Convox, ECS and Docker for most of our backend services, but this solution isn't a perfect fit for all our use cases.

    Read more
  • Service to Service Authentication with Paladin

    A few weeks ago, we had a hacking session at Opendoor. As we continue growing, we're starting to implement an increasing number of smaller, more focused services within our stack in a range of languages. For my hack week project I decided to work on a way to authenticate these services both inside and outside of our private network. Service authentication becomes increasingly painful as the number of services you support grows.

    Read more
  • Serving Analytics the Right Way

    Every company uses some kind of database to store its data, and in order to survive and thrive, it is crucial to use this data effectively in decision making. However, getting value from our data is a lot easier said than done. In this blog post, we'll discuss how Opendoor takes a modern approach to serving analytics, making sure that everyone on the team gets the data they need to make high-quality decisions.

    Read more
  • Iterative Model Development

    We recently hosted the SF Bay Area Machine Learning meetup where we discussed our approach of iterative model development. Here's the talk excerpt and the slides.

    Read more
  • Producing Error Estimates with Residual Modeling

    In a previous post, we discussed the need for understanding the confidence of our housing price predictions. One strategy for doing this is building a confidence model. In this post, we will describe one possible solution and the way we use it at Opendoor. Specifically, we will describe an error model \( \hat{g} \) that estimates the prediction errors from a housing valuation model \( \hat{f} \) that predicts housing prices.

    Read more
  • Quantifying Uncertainty

    Opendoor provides fair market offers to homeowners so they can sell their homes to us with confidence. In this post, we'll discuss how understanding the prediction error curve of our algorithms has guided our valuation process for these offers.

    Read more
  • Moving Opendoor's Data Science stack from Heroku to Convox

    At Opendoor, we've run our apps on Heroku's platform since our first deploy. We're big believers in the platform's value proposition: get up and running quickly and avoid spending time dealing with infrastructure.

    Read more
  • Phoenix on Rails for Client Push Notifications

    At Opendoor, we're growing our application to take the pain out of Real Estate. Like most places we have a front end that our customers interact with, and a back-end system that our operations staff use.

    Read more
  • How homeowners can identify hot neighborhoods before they pop

    The neighborhood you choose to buy a home in may be the most significant investment decision you will ever make. Buy the right home, in the right neighborhood, at the right time and your net worth can skyrocket.

    Read on medium