Abstract—Process mining has emerged as a powerful approach for analyzing business processes through event logs. Still, it often requires specialized expertise to interpret results and lacks autonomous decision-making capabilities for implementing improvements. This paper presents a novel framework integrating process mining with agentic artificial intelligence to create an autonomous system for continuous process discovery, analysis, and optimization. We introduce a specialized multi-agent architecture where five distinct agents collaborate through sequential knowledge transfer using the CrewAI framework: the Data Processing Agent performs event log preprocessing and quality assessment using PM4Py format validation with schema compliance checking; the Process Analysis Agent discovers process models through inductive mining algorithms and directly-follows graph analysis; the Workflow Pattern Agent identifies process variants and conformance patterns using statistical significance testing; the Bottleneck Detection Agent combines domain knowledge with regulatory requirements for anomaly detection; and the Process Optimization Agent generates quantifiable improvement recommendations through simulation-based impact analysis and ROI calculations. The framework implements sequential task execution with context dependency chaining to prevent analytical overlap while ensuring systematic knowledge transfer between agents. Our experimental validation on a real-world customer support dataset of 8,469 tickets demonstrates that our approach achieves 98.7% accuracy in data preprocessing, generates process models with 0.94 fitness and 0.86 precision scores, identifies bottlenecks with 92.3% recall, and produces optimization recommendations resulting in 23.7% reduction in process cycle time, 18.4% improvement in resource utilization, and 15.9% estimated cost reduction. Our contribution bridges the gap between technical process mining capabilities and business-oriented decision-making, making process optimization more accessible to non-experts while delivering measurable business value.
 
Keywords—agentic Artificial Intelligence (AI), process mining, process optimization, business process, optimization, agents, Artificial Intelligence (AI)


Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).