Learning from successful urgent automation of case processing of public services during COVID-19