Best Practices for Human Evaluation in Phoenix Experiments

Hi, I've been running some experiments on Phoenix, and things have been going well. I'm now looking into best practices for incorporating human evaluation for a specific metric. Is it possible for a human reviewer to log into the Phoenix platform and directly add their evaluations there? Or would it be better to collect the evaluations through another tool and then import the results into Phoenix? Would love to hear your suggestions on this matter 😃