This was resolved by dropping all of the DB tables. I’m assuming that the migrations were interrupted during a deployment.