@Victor_Akolawole for a big branchy workflow the biggest win is not re-running the whole thing every time. if youre chasing an actual failure, go to the Executions tab, open the failed run and hit Debug in editor, that pulls that exact runs data back onto the canvas pinned on the first node so you can replay the real failure instead of guessing at it. then to find which branch is off, select a node and use Execute step, it runs just that node and whatever feeds into it using the same input, so you can poke one branch at a time without firing the rest. pin the data on your trigger/source nodes too so youre not re-hitting external systems on every test, and temporarily disable the branches youre not looking at to shrink the surface. are you debugging a hard error, or a branch quietly returning the wrong data or taking the wrong path? the approach shifts a bit between those.