Luke Kenworthy

2021 - Present, US Space Force
Test Analyst

Lead data scientist and test analyst for the GPS enterprise operational test team. Ensured statistical rigor to the testing of $21B+ of satellite systems. Other efforts include:

• Won BRAVO 0, the first-ever classified DoD hackathon (90+ participants), where we built mission planning software that was lauded by the Air Force Vice Chief of Staff.
• Created and taught in Vault University, a series of course taken by 300+ people across four services how to use cloud computing to operationalize their data.
• Completed four-month Air Force-MIT AI Accelerator fellowship, where I researched AI applications for pilot fatigue detection.

2019 (Summer), GoodTime
Software Engineering Intern

Worked in a full JavaScript stack (Node, Express, React, Redux) to add new features to the GoodTime's premier interview scheduling app. Received a return offer.

• Built first visitor management system integration to provide a seamless candidate experience for in-person interviews.
• Enhanced user navigation by developing mobile navigation bar, ensuring seamless functionality.

NICE: Robust Scheduling through RL-Guided Integer Programming

Luke Kenworthy, Siddharth Nayak, Christopher Chin, Hamsa Balakrishnan

Association for the Advancement of Artificial Intelligence (AAAI) 2022

In which we define a new method, NICE, which uses reinforcement learning to guide the formation of integer linear programs. We use this method to effeciently build robust flight schedules.

Fatigue Assessment from Facial Videos using Deep Neural Networks and Engineered Features Informed by Domain Knowledge

Luke Kenworthy*, Patrick Moore*, Hrishikesh Rao, Laura Brattain, Kevin James, Thomas Heldt (*Equal contribution from authors)

IEEE Engineering in Medicine and Biology Society (EMBC) 2023 (publication pending)

In which we build two methods of efficiently using machine learning for detecting fatigue from facial video and compare them.