Autonomous experimentation framework that goes beyond basic A/B testing. Use when: run experiment on [channel], check for conflicting experiments, score this test, what’s in the playbook for [channel]?, calculate sample size for [metric] with [mde], switch to bandit mode.
