There’s a subtle SDK behavior difference between bridged and blind transfers that isn’t well documented.
In the Python SDK v2.5.1, calling post_conversations_call_participant_replace for a blind transfer returns a 202 Accepted, but calling it for a bridged (consult) transfer returns a 200 OK with the new conversation ID. I documented this discrepancy in Issue #428 on the Genesys GitHub.
From a project management perspective, choosing the wrong transfer type during migration can introduce significant risk.
Our PureConnect environment exclusively used consult transfers because agents needed to brief the receiving party. If we accidentally deploy blind transfers in GC, the receiving agent gets a cold call with zero context, leading to customer frustration and a spike in repeat calls.