Historically, the mortgage industry hasn’t embraced technology. Many tasks are accomplished via email and phone calls. When borrowers, especially first-time homebuyers, have questions about their loans and the mortgage process, they often have to get on the phone or examine a spreadsheet sent by a loan officer.
At Better Mortgage, we’re creating software that brings transparency and speed to the mortgage process. Borrowers can look up their rates, see how much they can afford to borrow, and compare loan options—all from a convenient online dashboard. Better Mortgage also aims to decrease loan processing time and costs so borrowers can get a faster and cheaper mortgage.
The Better family of companies has a vision to help homeowners through the entire homeownership process, through its affiliates it can help with purchase, refinance, real estate, and insurance.
How do we do it?
Our product and engineering teams have a relentless focus on fast iteration and experimentation. We hire engineers by focusing our interviews on day-to-day problem solving skills and analytical capabilities rather than algorithm exercises and puzzles. We’re always looking for ways to improve customer experience and efficiency of our loan processing teams. Whenever there is an idea or hypothesis that needs to be tested, we’re able to rapidly build a proof-of-concept and A/B test it to validate the hypothesis. This allows us to quickly improve our product in ways that traditional lenders can’t.
For example, each day, our loan expert teams work with hundreds of borrowers. To facilitate communication and quickly gather context, we integrate with all our communication providers, ingest data in real-time with a webhook, and then sanitize and attribute the data to the borrower to give our teams quick access to a borrower’s loan file, speeding up productivity and increasing efficiency.
Tech at Better
Better’s infrastructure is built on Amazon Web Services (AWS) and is made up of Kubernetes clusters containing many microservices written in NodeJS, Python, and Go. In 2017, Kubernetes was still an up-and-coming technology, but the engineering team at Better had already deployed a Kubernetes cluster in our production environment and installed all their applications onto it, bridging the gap between engineers and infrastructure without needing to rely on a DevOps engineer.
We ship code over 100 times everyday thanks to our comprehensive, automated Continuous Integration and deployment (CI/CD) pipelines. The Better engineering team was built on the idea of fast and iterative development, meaning that product managers and engineers need the ability to quickly develop, test, and ship features or changes into production.
Our infrastructure is heavily instrumented with Datadog. We can get highly detailed performance insights into our servers, traffic, code, and more. These insights allow us to build robust monitoring and to find and fix scalability issues easily. It can also give very detailed insights into server and browser application logic, allowing us to identify performance bottlenecks and helps us scale our software as the business grows.
My role as an engineering manager at Better
As engineering manager, I am responsible for ensuring the growth team can plan and implement any features and experiments at the top of the funnel. I also mentor and help engineers on my team develop and progress in their careers.
About 40% of my day is spent working with our amazing product team and engineering teams to design and plan our roadmap. And another 40% is spent working with my engineering team to help with questions, challenges, and career development.
The rest is hands-on coding, learning, and research.
The great teams and the innovative tech make Better a really great place to work. Interested in joining us? Check out our career opportunities.
- tech team
Upserts in RedshiftRedshift doesn't support upserts (updates + inserts) but using a few tricks we can implement it anyway.Wed Aug 28 2019—by Erik Bernhardsson1 min read
Tackling Conversion With Survival Models And Neural NetworksAt Better, we have spent a lot of effort modeling conversion rates using Kaplan-Meier and gamma distributions. We recently released convoys, a Python package to fit these models.Thu Aug 26 2021—by Marshal Bratten, Jesse Lax, and Stephen Ma3 min read
Engineering a Diverse Team: Taffy Chen and Jimmy FarilloSoftware engineers Taffy Chen and Jimmy Farillo launch a new blog series to showcase different perspectives on the Better engineering team, and the ways they’re working to make it even more diverse and inclusive.Fri May 14 2021—by Taffy Chen and Jimmy Farillo2 min read