Public Utilities
SAP demonstrates AI agents for utility workflows
SAP showcased AI agents automating utility back-office workflows while emphasizing clean data and human approval for exceptions and safety-sensitive decisions.
SAP demonstrated an autonomous-utility scenario in which AI agents handled meter-to-cash exceptions and processed interconnection applications. In the example, agents worked according to business rules, processed about 800 applications and routed 11 exceptions to a program manager with recommended solutions. A human remained responsible for final approval and could intervene. Utility executives and panelists also stressed that data quality remains a prerequisite for successful AI deployment. Southwest Gas reportedly spent nearly five years restructuring its data, while other utilities discussed using AI to support field technicians without removing human oversight from safety-critical decisions. The demonstration shows how utilities are moving beyond experimentation toward workflow automation. It also underscores that implementation depends on clean data, domain expertise, review processes and carefully defined limits for autonomous action.
Original source: ASUG