Comment by arpa

Okay, this approach is really cool, but... just hear me out: what if... you don't store personally identifiable data in the event store? What if you only store references/ids that point to services which can resolve those ids to data, and that data needs not to be immutable. In the event of erasure request you shred/anonymize that data only, without touching kafka. I mean it's pretty obvious to me, to the point i feel like i'm missing some huge point and grossly misunderstanding the whole problem