How to setup prediction at week level in Mixed effect Models panel data

Question

I am doing a research study where I am trying to design a Methodology to design Mixed effect regression models. I have sample CPG panel level data as below -

enter image description here

Can you please give an idea How to build a Mixed effect prediction model on this data? My initial thoughts were to have fixed effects at overall data, and random effects group at PPG/product group level, however how to make weekly predictions in this case, or in case its not necessary to make weekly predictions and just making general prediction ?

Thanks for help in advance.