Applying generative AI to reduce administrative burden for WMO consultants

Abstract

Social welfare support consultants have to deal with recurring administrativeburdens, which can cost them a large portion of their time. Social welfaresupport consultants, working under the Dutch law of social assistance, assessa citizen’s situation to provide possible solutions to ensure their participationin society. During the home visit of the citizen and later, the accompanyingadministrative burden can require a significant amount of attention from theconsultant. This case study aims to assess possible ways of using generativeAI (GenAI) to reduce this administrative burden.Two ways of reducing administrative burden were explored, throughthe introduction of a system. During a conversation, the consultant has totake notes; however, the AI-assisted system could be used to record, transcribe, and summarize the conversation, so only minimal notes would have tobe taken. This ensured that the consultant could pay more attention to theconversation itself. After a conversation, the consultant could load the AIgenerated summary, only requiring the consultant to make alterations ratherthan having to write a summary from scratch. The system was tested on asmall group of participants, where results indicated a significant decrease inperceived administrative burden and administration time.