The increasing complexity of mobile networks as they evolve towards 6 th Generation (6G), coupled with the importance of providing fast responses to performance degradation issues, requires the adoption of cost-effective Trouble-Ticket (TT) management systems by Mobile Network Operators (MNOs). These systems are pivotal in ensuring compliance with Service Level Agreements (SLAs) and maintaining Quality of Service (QoS) standards. In this context, the introduction of a new Smart Operations (SO) framework referred to as Smart Trouble Ticket Management (STTM), promotes the integration of Artificial Intelligence (AI)/Machine Learning (ML) to leverage performance data, aiming at early performance degradation detection and an efficient handling of TTs. This paper presents a novel methodology for STTM that relies on a ML clustering algorithm to identify performance degradation instances when cells transition to lower-ranked clusters. The benefits of this dynamic approach are twofold. Firstly, it offers an automated and adaptable process capable of accommodating diverse Key Performance Indicators (KPIs). Secondly, it enhances the reliability of TT opening and closure compared to conventional approaches, particularly those reliant on fixed threshold values. The proposed methodology, tested on live network data, achieved an improvement of 5.53% in TT Opening Accuracy Rate (OAR), and reduced the TT Repeat Call Rate (RCR) by 17.57%.