>> The best fix imo is to slather a battery of views over your data to minimise the risk of getting the joins wrong, and it'd be interesting to see how that app

>> The best fix imo is to slather a battery of views over your data to minimise the risk of getting the joins wrong, and it'd be interesting to see how that approach could improve the bot's quality. Yes, the success of the bot's query (at least on GPT-3) is a combination of (a literal and good) prompt and how well did you explain each of the columns to the model.