Josh Reini

Developer Advocate and AI Engineer.

I help developers build AI systems they can trust—through open source tooling and technical education.

What I'm working on

AI agents that actually work. Most agents fail silently. I build evaluation frameworks and observability tools to catch failures before users do. My GPA framework gives teams a practical way to measure agent reliability.

Skills for coding agents. Building skills to make agents more reliable with the help of coding agents like Claude Code and Cortex Code.

Education at scale. I've developed courses on DeepLearning.ai, Coursera reaching 40k+ engineers. I also advise Stanford CS329T: Trustworthy Machine Learning.

Open Source

Co-creator and maintainer of TruLens, the open-source library for evaluating and tracing AI agents.

Interests beyond AI

I'm interested in economics, monetary policy, and market structures, and sometimes write on the subject.