synthetized dataset based on documentation not publicly available q&a pairs in json format (for now).