The likely cause for this issue is crosstalk. This usually happens when the guest isn't wearing headphones. This allows your voice to leak back into their microphone which will result in a track that has multiple voices on it.
This will show as an echo effect in the postproduction. Unfortunately, there isn't an easy fix aside from editing the track and replacing the unwanted voice with silence. You can load your solo tracks into an audio editor and cover the unwanted voices on each track with silence then you can mix them together.
If the problem still persists, contact Zencastr support vial email or chat.