Hi HN,

We are trying to make different Agent Memory systems comparable without letting each team choose its own answer model and evaluation pipeline.