[RESOLVED] Issue was localized to a specific set of user-roles and limited to one of the API servers. Issue is fixed now.
We are experiencing higher than normal API latency on one of the database segments and we are working on it.
