Short, dated observations from what I'm building.
The hard part of production text-to-SQL is deciding what fits in the prompt before generation starts. In DeepQuery that’s picking at most six tables for each question and sending only those, not the whole schema. Most of the engineering effort goes into context management.
Every time someone dismisses planner/worker as “just a ReAct loop,” they miss the part that actually matters: the planner is where you bound failure. Workers should be allowed to fail loudly; the planner is what decides whether that failure matters. Without that boundary, multi-agent systems collapse into a spaghetti of error-handling.