There's a stream lib (not included in core, it's a separate package), which fulfils some data flow concerns. The DbMonster implementation could have used those. TBH using Redux for DbMonster is a bit besides the point of the exercise IMO.