Blog

Thoughts on ML systems, AI product building, and lessons learned from shipping to millions.

Coming Soon

I'm working on a few posts right now. In the meantime, here's what's on deck:

🏗️

Deploying ML at National Scale

Lessons from shipping Health Sentinel to 400M+ users — data drift, alert fatigue, and monitoring in the wild.

🧪

Building AI Products That People Actually Use

How Pulp AI went from a weekend hack to a simulation engine product teams rely on daily.

⚠️

Real-World Systems Failures

Post-mortems and pattern recognition from production ML incidents I've seen (and caused).

Want to be notified when the first post drops?

Drop me an email →