Production engineering across AI/ML, backend, and platform work.
Software Engineer · Nov 2025 – Present
NovumAI is the first AI system I've shipped that has to feel instant — coaching suggestions that fire while a sales agent is on a live call. It's also the kind of work I picked on purpose: small team, primary AI engineer, end to end. The hard part of real-time coaching isn't the model. It's knowing when the model is actually right.
Visit Site →
Software Engineer · Nov 2022 – Mar 2023
Engagebud was my first full-time job, five months of product engineering before I left for grad school. It was also the first time my work touched real money — actual Shopify storefronts, real conversion numbers I had to defend. I built a Thompson Sampling bandit for campaign routing, and the honest answer about ML metrics at small startups is that you rarely get clean attribution — that's a thing I think about a lot now.
Visit Site →
Claude Code marketplace plugin
Claude Professor is a Claude Code plugin I built because I was the user it was for. I'd been using AI to write production code and catching myself writing things I didn't fully understand. The plugin is structural: it gates design sessions behind concept teaching using a state machine where the wrong path isn't possible, not just discouraged. The point is to leave the session understanding what you wrote, not just having written something.
View on GitHub →
Multi-tenant Slack volunteer onboarding bot
Sherpa started as an internal Slack bot I built at Changing The Present to onboard new interns. After it worked there, I generalized it into a distributable platform any nonprofit can install — multi-tenant, OAuth, per-workspace cost protection. It's a side project, not a startup, but production-grade engineering doesn't require production-grade scale to be worth doing. Most of what I learned about platform work came from building something usable that nobody is using.
View on GitHub →
Live auction bidding system
FlashBids was a six-person academic group project, and I owned two pieces of it: the bid processing pipeline and the streaming leaderboard. It was the first time I had to worry about ordering and race conditions under live load — 70 people bidding simultaneously in the classroom pilot, with sub-200ms updates the whole way through. SQS FIFO did most of the work for ordering, but figuring out which guarantees the architecture actually needed (versus which ones felt safer to claim) was its own exercise.
View on GitHub →
Roles where the work was internal tooling — no public artifact to link, but worth naming.
Software Engineer Intern · Jul – Oct 2025
Built the AI infrastructure for a volunteer-onboarding platform: a Slack onboarding agent (later generalized into Sherpa) that cut intern ramp time from about 5 days to 2 across 50 hires, a Jotform-to-RDS ingestion pipeline handling 10K+ records, and a Retool HR dashboard with bidirectional Asana sync for the 100-200 interns being managed at any time.
Software Engineer Intern · Jan – Jun 2022
ML infrastructure for predictive device-failure detection on the Mist AI team. Built PySpark telemetry ETL with schema validation (15% recall improvement on early failure detection), PSI-based feature drift monitoring with automated retraining triggers (20% reduction in false-positive alerts), and a dual-mode Junos config tool in Perl/LISP that cut test failures by 30%.