I like building quality backend systems that enable impactful missions. I most recently worked at Robinhood building ML-based fraud defenses. I'm particularly interested in generative AI and its potential to shape the future.
As a member of the anti-fraud team, I developed ACH transfer fraud defenses in Python. I worked closely with a data science group to launch real-time machine learning inference pipelines to identify fraudsters and implemented targeted account restrictions to protect the firm from fraud loss while enabling a positive user experience.
•
Implemented new fraud evaluation flows, generating a total of $1 million in monthly fraud savings
•
Created 100 real-time and offline machine learning features for 20M users
•
Worked with two engineers to ship a cash advance product, resulting in a 10% increase in early access funds with no change in fraud loss
I worked on an integrated hardware/software project in the Autonomous Systems division's Maritime Undersea program. I developed the full stack of a Java application that commands and monitors a remote distributed computing system.
•
Wrote a Python code generator that translates C++ network message type definitions into Java serializer classes, saving 500 developer hours
•
Implemented a modular message processing system that reduced the time to onboard new message types by 50%
