Hands-On Lab: Optimize Cost and Latency

This lesson is for enrolled learners

The concept This lab is a constrained optimization. You will take a feature that already has a quality gate and an informal or written cost or latency budget, and you will try a small set of changes one at a time: cache or prefix reuse, a cheaper model on an easy slice, a shorter prompt, fewer tool calls, a deferred……

Enroll to continue Already enrolled? Sign in