Machine learning, software, and data engineering.
M.Eng. in Artificial Intelligence at Duke University.
May 2026 — Present
Unlimited.ai
December 2025 — Present
Guided
October 2025 — Present
Duke Applied AI Cyber Lab
February 2024 — April 2026
Education Credit Union
May 2026
A DreamerV3 agent that recovers a Boeing 737-300 from stalls and upsets. As far as the literature shows, the first time a world-model approach has been used for stall recovery.
March 2026
A defense stock claim analyzer that scrapes tweets, filters bots, and labels claims as exaggerated, accurate, or understated by comparing them against actual price movements.
November 2025
A neural network-powered risk authentication system for Shibboleth IdP that learns each user's unique login behavior to detect account compromise in real-time.
February 2026
A behavioral benchmark that evaluates whether AI systems know when to stop answering and start asking in high-risk situations.
Notes on machine learning, infrastructure, and building things that hold up.
November 17, 2025
Two-thirds of college students feel lonely. At Duke's AI Hackathon, we built Sync AI to change that.
November 4, 2025
You log into Canvas, your email, and your university portal every day without thinking twice. But the security protecting those logins? It's nowhere near as sophisticated as it should be. So my team and I decided to modernize it.
October 1, 2025
A data storytelling project exploring whether America's growing network of data centers is truly straining the power grid, or if the story in the numbers points to something more complex.
May 24, 2025
I self-host my applications. The biggest downside? Downtime during updates and reboots. Here's how I load balanced my entire stack to keep things running, even when a server goes down.
M.Eng. in Artificial Intelligence at Duke University
Hi! I'm Sam. I'm an engineer at Duke pursuing an M.Eng. in Artificial Intelligence. I've been writing software since 2013, when I started as a kid with Minecraft plugins, websites for nonprofits, and Arduino projects, and these days most of my work is data and machine learning systems.
My past technical experiences include a distributed reporting system for a credit union, risk-based authentication for Shibboleth, and a reinforcement learning agent that recovers a Boeing 737 from a stall. I was previously a Teaching Assistant for AI Entrepreneurship at Duke's Fuqua School of Business. Right now I'm a Software Engineering Intern at Unlimited.ai, where the work is mostly Databricks, data pipelines, and large system design. I'm also the Chief Technology Officer at Guided, an ed-tech startup, and a Research Fellow in the Duke Applied AI Cyber Lab, where I build models that detect encrypted VPN traffic slipping past institutional web filters. I'm broadly interested in machine learning systems, data engineering, authentication, and infrastructure.
Outside of work and school, I try to spend my free time away from screens. I like being out in nature and going on hikes, and I take a lot of photos of the places I travel to, along with more of Duke Gardens and campus than I can count. I also collect vinyl, and I enjoy messing with EQs and audio mixes, mostly for my own ears. And yes, this website runs on servers I manage myself.
education
2025 — Present
Duke University
GPA 3.96 / 4.00
2021 — 2025
West Texas A&M University
GPA 3.94 / 4.00, with Beta Gamma Sigma, Upsilon Pi Epsilon, and the 2025 Carey Lyles Emerging Technology Award
I'm always happy to talk about machine learning systems, data engineering, or something you're building. Send me a message and I'll get back to you shortly.
