Production of value-added, plant-derived compounds in microbes increasingly attracts commercially interest in food and pharmaceutical industries. However, plant metabolic pathways are complex, require a robust balance of enzymes, cofactors, ATP and other metabolites, and often result in low production when transplanted to bacteria. This is exemplified by the biosynthesis of curcuminoids from theCurcuma longaplant. Here, we combine dynamic pathway modeling, systematic testing of isoenzymes, and the optimization of gene expression levels and substrate concentrations for the biosynthesis of curcuminoids inPseudomonas putida, leading to unprecedented conversion rates of caffeic acid and tyrosine to curcumin. The development of kinetic ensemble models guided the design of production strains, emphasizing the necessity of high relative expression ofc3h, curs2anddcsand, the low relative expression oftal, comt, ccoaomt, and4cl4. This optimization resulted in a strain that achieved a 10.8 ±1.8% of the maximum theoretical yield of curcumin from tyrosine. This represents a 4.1-fold increase in production efficiency and the highest yield reported to date, demonstrating the potential ofP. putidaas a promising platform for curcuminoid production. Our findings highlight the effectiveness of our strategy not only in the advances in the production of curcuminoids but also in setting a framework for the biosynthesis of other complex compounds.