One of the downsides of okrs is that they can require significant human effort to make the check of logs and metrics to see if the key result is being met.

Has anyone had success with using AI to reduce this burden?