⁠if we switched models today then how do we monitor response quality qualitatively and quantitatively?

Currently we use gemini 2.0 flash for answering, if we were to switch models in the future then how we will not know how well the new model is performing because we arent analyzing response quality. The only analysis we do is when someone from the team takes a manual look at the dashboard. - Research ways of how other companies do this - This should be a solved problem so what tools are available for this? First connect with Siddhant to discuss the found methods/tools before implementing anything