Tim Hopper

Roadblock title:

How to use streaming machine learning for real-time data

Time:

Thursday - 3:00 PM (4th floor - Indigo)

Abstract:

Varo Bank is a new kind of bank ⎼ the first nationally-chartered consumer techbank in the U.S., built from the ground up with a focus on the needs of Americans living paycheck to paycheck. From credit building to savings to faster payments, Varo has a complete solution to help hard-working Americans progress financially.

At Varo, we want to embed data and machine learning across all products and consumer experiences, requiring real-time data derived from multiple sources. We considered developing a new backend microservice to compute and serve data to address this challenge.

Instead, we decided to reuse its existing streaming machine learning feature platform to compute and serve this data. This talk demonstrates how the bank leveraged its existing technology to address this challenge.

Bio:

Tim Hopper is a Staff Software Engineer on Varo Bank’s ML platform team. He has over ten years of experience working with Python and machine learning in cybersecurity, weather forecasting, geospatial modeling, and public health research applications.