With emergence of nanotechnology, it is possible to control interfaces and flow at nanoscale. This is of particular interest in the Oil and Gas industry (O&G), where nanoscience can be applied on processes such as Enhance Oil Recovery (EOR) and oil mitigation. On this direction, one of potential strategies is the so called Nano-EOR based on surface drive flow, where mobilization of hydrocarbons trapped at the pore scale can be favored by controlling by the chemical environment through “wettability modifiers”, such as functionalized nanoparticles (NP) and surfactants. The challenge consists then to search for optimal functionalized NP for oil recovery and mitigation at the harsh conditions found in oil reservoirs. Here, we introduce a hierarchical computational protocol based on the role of NP interfacial and wetting properties within oil/brine/rock interfaces to the fluid displacement in pore network models (PNMs). This integrated multiscale computational protocol ranges from first principles calculations, to determine and benchmark interatomic potentials, which are coupled with molecular dynamics (MD) to characterize the descriptors (interfacial properties and viscosity). The MD results are then mapped into Lattice Boltzmann method (LBM) simulation parameters to model the oil displacement process in PNMs at the microscale. Here, we show that this multiscale protocol coupled with Machine Learning techniques can be a resourceful tool to explore the potentialities of chemical additives, such as NP and surfactants, for the oil recovery process and investigate the effects of interfacial tension and wetting properties on the fluid behavior at both nano and microscales.