RoDia introduces the first dataset for Romanian dialect identification from speech, aiming to stimulate research in this low-resource language.