I have never done it before but I it is feasible. Here is one paper about genomic screen based on Crispr/Cas 9. In the schematic figure, namely figure 1. They constructed the stable-expression cellline which was transducted by 3 three different plasmids: CCR5 gene,  LTR gene and Cas 9 cell in the CEM cell. Therefore, it may also work in your situation.

If using a lentiviral overexpression technique, it depends on the size of the construct. If the genes are small enough, you can overexpress them by driving them with a constitutively active promoter and separate them with a self cleaving peptide thereby expressing them in a 1:1:1 ratio. 

Thank you all so much for your answers. Will try to do as soon as possible.