Shorter Improvement Lead Times and Field-Level Control Are Key to AICC Success, Says MindwareWorks CAO

At the SWCC 2026 conference held at COEX in Seoul on September 9, MindwareWorks CAO Jaein Lee explained that the success of an AI contact center depends not simply on adopting generative AI or AI agent technologies, but on establishing a system that can quickly identify failures in real customer interactions, improve them, and reflect the changes in operations.

MindwareWorks presented an operational loop that collects failed interaction cases, analyzes their causes, develops improvement measures, and validates and deploys those changes. The approach continuously examines issues that arise in the field—including speech-recognition errors, changes in customer intent, unexpected questions, and system-integration delays—and incorporates the findings into scenario improvements and testing.

The presentation also introduced a direction for automating these validation and improvement processes with AI-agent-based systems. By analyzing failed voice and conversation data, generating improvement proposals, and applying regression and adversarial testing before deployment, the company aims to shorten improvement lead times and enable continuous operational enhancement.

For mission-critical areas such as financial services, where financial and legal risks are involved, MindwareWorks emphasized the need to clearly define the role of generative AI while retaining validated rule-based logic and control by field experts. The company plans to focus on providing an AICC operating environment in which field organizations can directly manage the process from implementation and deployment through daily improvement.