Sistem Agen AI (Agentic AI) sebagai Mitra Produktivitas UMKM di Indonesia

Abstract

MSMEs are the backbone of the Indonesian economy, yet their productivity remains hampered by limited resources, digital readiness, and business process governance. The emergence of agentic AI, artificial intelligence-based agents capable of autonomously planning, remembering, and executing tasks, offers opportunities to reduce administrative burdens and improve workflow efficiency. This study aims to map the opportunities and barriers to implementing agentic AI in MSMEs, design an Agentic-MSME Alignment (AUA) framework, and examine its impact on productivity and service quality indicators. A mixed-methods approach was used, including surveys of 30 MSMEs in three major cities, in-depth interviews with 10 MSMEs, observations of AI prototype implementations, and system log analysis. Data were analyzed using paired sample t-tests and thematic analysis to validate the results. The findings show significant improvements in time efficiency, sales, and customer satisfaction. Interviews confirmed that the greatest benefits were accelerated customer response and reduced administrative burden, although concerns were raised about data privacy and output errors. The integration of the AUA framework has been shown to strengthen transparency, reduce administrative errors, and encourage a new division of labor between AI and humans. This study provides empirical evidence that agentic AI can be a productivity partner for MSMEs, provided there is increased digital literacy, ethical governance, and community support. This research contribution not only broadens academic knowledge on AI in MSMEs but also offers practical guidance for developing inclusive digital transformation policies and strategies in Indonesia.