Zheng Liu, Institute for Advanced Study, Tsinghua University Density functional theory (DFT) is practically a static mean-field theory, and it is commonly perceived that the application of this method to correlated systems is severely restricted. Nevertheless, this talk aims to show that by carefully designing the calculation, DFT can still provide important physical insights. We will illustrate this perspective based on two examples: (1) nematicity in FeSe; and (2) Mott physics in 1T-TaS2. We will identify novel properties calculable by DFT, and benchmark the calculation results with experiments. These calculations in addition suggest hidden orders in FeSe and giant renormalization of the correlation strength in 1T-TaS2 by lattice vibration, which await experimental justification.