I realize this is a bit on the late side, but we are opportunistically looking for someone with machine learning and computer vision background to come work with us for a few months during the summer of 2018.
The problem we are trying to solve is OCR of semi-structured data, typically documents like tax returns, paystubs, bank statements, and other similar documents. There are great off-the-shelf tools (such as Tesseract) for extracting text from images, but this is just one part of the problem. We also need to take the extracted data (typically numbers) and reconcile them with various fields we are trying to find. For instance, paystubs are highly nonstandard and come in many layouts and shapes, but they generally contain the same fields such as gross pay or year to date state tax.
We have a lot of document data to work with, and some (but limited) “ground truth” data. It is possible that a substantial component of this job is to build internal tools to collect more training data.
The ideal candidate has experience with Python and computer vision, ideally with tools such as Tesseract, OpenCV, Keras, or Scikit-Learn. You do not necessarily need experience with all of them, though. You also have some experience building computer vision / OCR / machine learning systems. We also want to see some experience using Python as a general-purpose tool outside of machine learning.
This is a paid internship and our office is in Soho, New York. If you are interested, feel free to email erik at better dot com with a resume and we will get back to you quickly!
About the company
It’s hard not to be excited about mortgages when you’re actively upending the conventions of a $13 trillion dollar business that impacts over 200 million Americans.
Rife with antiquated processes, the home finance industry has remained fundamentally unchanged since the 1970s. At Better, we’re building a next-generation mortgage lender from the ground up. Our team combines leading software engineers from tech giants like Google and Spotify with financial talent from firms like Blackstone and Bridgewater, creating a hybrid institution unlike any other.
Help us hack a thirteen trillion dollar industry by building a product that will allow more people than the status quo to own a home and build wealth rather than rent for life. Our tech team is small, and you will be a big part of defining the technical direction and culture. We encourage proposals for projects off the beaten path, experimentation with different frameworks and libraries, and doing as you see fit to solve problems. We also offer above-market compensation and equity, as well as and full benefits.
For more job opportunities, see our job board!
- machine learning
Using ImageMagick for PDF redactionAutomating document redaction with a classic command line toolThu Aug 06 2020—by Robert Brownstein3 min read
Modeling conversion rates and saving millions of dollars using Kaplan-Meier and gamma distributionsAt 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.Mon Jul 29 2019—by Erik Bernhardsson8 min read
Experimenting With Daily Email - Which One Performs Better?Mortgage rates move daily, and we like to keep our customers aware of them, but does awareness drive conversion?Wed Aug 19 2020—by Brent Wisener5 min read