Glycyrrhetinic acid 3-O-monoglucuronide (GAMG) and glycyrrhizin (GL) are typical pentacyclic triterpenoid saponins present in licorice roots that exhibit important pharmacological properties. GAMG and GL are sourced primarily from licorice, namely, "digging roots and extracting acid", which results in ecological imbalance and environmental pollution. To overcome these issues, it is an eco-friendly and sustainable alternative to de novo synthesize GAMG and GL in Saccharomyces cerevisiae by coupling UDP-glucuronic acid and glycyrrhetinic acid (GA) under the catalysis of glycosyltransferases. However, the reported productivity is too low for industrial production. The primary limiting factors include insufficient supply of precursors, poor activity, and low specificity of the glycosyltransferases. In this study, we reconstructed the GA-producing platform strain in which GA was mainly accumulated intracellularly, making it suitable for subsequent glycosylation modifications. Different glycosyltransferases were tested and selected in the platform strain for achieving the efficient and specific synthesis of GAMG and GL. Protein engineering of rate-limiting enzyme cUGT73P12 was performed to further facilitate the conversion from GAMG to GL. The glycosylation capacity of this system was assessed in vivo, uncovering that the insufficient supply of precursor GA limited glycosylation modifications. Correspondingly, multiple metabolic engineering strategies were applied to optimize the carbon flux distribution, increasing the titers of GAMG and GL by 115.6 and 50.2%, respectively. Finally, 552.9 mg/L GAMG and 476.6 mg/L GL were produced in the 5 L fed-batch fermentation, which were 238and 79-fold higher than previously reported values, respectively. The strategy described herein can serve as a reference for the glycosylation of triterpenoids in a cost-effective manner.