Are these commercial batteries or batteries that you made? If you make the batteries in lab you need to allow sufficient time for the electrolyte to fully saturate the cathode (I wait 12 hours after assembly before testing but shorter probably works). If electrolyte reaches new untouched parts of the cathode or anode during charging then voltage could drop.
Your y-axis is OCV, does that mean you are charging then intermittently measuring OCV? Open circuit voltage will decrease slightly after the end of a galvanostatic charge. Your plot looks like a dotted line, not data points? If you are doing intermittent galvanostatic charging then measuring OCV, the OCV measurement could be at a lower voltage than what galvanostatic charging ended at, depending on how much time has passed between the measurement. If possible you should just measure voltage during the galvanostatic test.