Engineering Agents is really UX Engineering
Building trustworthy agents is mostly a UX and state-sync problem, not just a framework or prompting problem.
Read →Tal Perry / Inference / Applied ML
Megawatts to Megabucks.
Recent writing
View all 30 →I Needed One Obscure Vision Model. Modal Made It Everyone Else’s Problem.
03The AI-Powered 10-Minute Habit That Taught My Kid to Read (And Made Me a Better Dad)
04Five Practical Lessons for Serving Models with Triton Inference Server
One-line biography: I've been writing about AI since we called it ML.
More about me →