José Carlos Sanz Soriano

The overload of modern IT Service Management (ITSM) Ticket and IT service management is facing a critical moment. Today’s companies operate in multicloud architectures, with distributed applications, hybrid environments, and a digital ecosystem expanding at an unsustainable pace. The consequences are felt where it hurts most: overloaded service desks, thousands of repetitive tickets, limited visibility into root causes, and increasing pressure to enhance the digital employee experience. Traditional IT Service Management (ITSM) models, based on static processes and manual rules, are no longer enough to keep up. Support teams are expected to absorb demand that grows faster than their capacity, while still trying to meet strict SLAs and avoid critical incidents in environments where every minute matters. In this context, AI applied to ITSM is no longer a nice-to-have, it becomes a strategic enabler: it allows us to anticipate failures, automate repetitive tasks, and enhance response capacity without increasing human effort. AI + advanced automation applied to the ticket lifecycle To transform ticket management, at Telefónica Tech we integrate a set of advanced capabilities based on AI, automation, and analytics. These technologies enable a complete redesign of the incident lifecycle, from creation to resolution, delivering efficiency and technical accuracy. Natural language processing (NLP) Through NLP models, users can create, update, or query tickets using natural language via chat, voice, or text. This reduces friction, eliminates categorization errors, and speeds up ticket intake. Autonomous agents (Agentic AI) Advanced automation brings in autonomous agents capable of: Learning from the service desk's historical data Predicting recurring incidents Recommending solutions Taking automatic actions or escalating cases based on impact or priority This enables a more proactive and self-sufficient service desk. Predictive Machine Learning Through statistical models and supervised algorithms, AI detects correlations within large volumes of operational data. This allows it to anticipate anomalous behavior and potential failures before they occur, triggering automatic alerts or remediations. Intelligent clustering and semantic analysis of tickets Semantic similarity algorithms help identify related incidents and cluster them around a common root cause. This speeds up the resolution of the whole group, reduces duplication, and enables early identification of incidents. These technical capabilities are the foundation of cognitive ITSM, enabling more agile, autonomous, and resilient operations. Cognitive ITSM in real-world environments At Telefónica Tech, we apply these technologies as part of our intelligent automation strategy in Business Applications, embedding AI directly into the customer’s daily operations. Our approach is built on three technical pillars: Intelligent ITSM architecture We design operating models that combine: Advanced automation AI applied to ticket flows Integration with multicloud/hybrid ecosystems End-to-end process orchestration Technology accelerators based on leading platforms We work with specialized platforms such as BMC Helix ITSM to enable features like NLP, semantic analysis, ticket clustering, or autonomous agents. These technologies are integrated as part of Telefónica Tech’s model, not as ends in themselves. Managed operation and continuous evolution Our differentiator lies not only in the technology, but in how we operate and evolve it. Deploying AI in ITSM is not a one-off project: it's a living system that requires ongoing supervision, fine-tuning, and continuous improvement. At Telefónica Tech, we address this complexity through a managed operations approach that spans the full lifecycle of the model: AI model tuning, as models are not stable indefinitely. Systems evolve, services change, user behavior shifts, and the technology ecosystem advances. That's why active model management is a core task: we monitor how ticket patterns evolve, detect gradual performance drops, and adjust the models when new incident types appear. —This ensures AI continues evolving and delivering value at the same pace as the business. Continuous analysis of ticket patterns, with ongoing monitoring of all generated tickets. It's not just about counting tickets, it's about understanding them. We analyze recurrence, load spikes, types of incidents, and their correlation with environmental changes, application deployments, or prior incidents. —This continuous tracking enables detection of structural issues, identification of anomalies, and insights into recurring incident areas, helping to anticipate service degradation and distinguish between short-term spikes and real trends. It’s a critical layer for transforming ITSM into a predictive, not just reactive, function. Progressive automation tuning, to adjust and optimize automation flows so they respond accurately and efficiently to real-world operational dynamics. Automated processes age too. A workflow defined six months ago may no longer be optimal if the application ecosystem, support model, or business priorities have changed. —That’s why we regularly review automation flows: we validate their efficiency, remove unnecessary automations, reinforce those that add value, and adapt others when they no longer fit operational reality. This ensures automation remains useful, sustainable, and aligned with the customer’s context. Technical observability applied to ITSM, providing continuous, detailed monitoring of IT systems and processes to understand behaviors. We integrate observability capabilities that correlate what's happening in services, how it impacts tickets, and how automation responds. —This helps detect anomalies before they become incidents, quickly analyze what went wrong, and continuously improve models based on real-world behaviors. It's key to anticipating issues and optimizing service management in automated and cognitive environments. Local and specialized support from our expert teams, offering direct assistance in the customer's environment and adapting to each organization’s specific needs (processes, priorities, and constraints) ensuring personalized and efficient incident and service management. —AI requires data, governance, and ongoing improvement. That’s precisely the value Telefónica Tech brings. Use cases Incident clustering and predictive problem management AI analyzes thousands of tickets to uncover hidden patterns across seemingly unrelated incidents. This approach enables the automatic grouping of related cases under a single root problem, preventing teams from working in isolation on repeated symptoms. —The result: a significant reduction in duplicated work and faster resolution of structural problems. Early detection of major incidents Models analyze, in real time, the volume, type, and speed of incoming tickets. When deviations from normal patterns are detected, the system can flag the potential onset of a major incident even before widespread impact occurs. —This enables proactive escalation and minimizes the impact on critical services. Automated solution recommendations Based on incident history, knowledge articles, and technical documentation, AI suggests actions with a high likelihood of success from the outset. —This significantly reduces mean time to resolution (MTTR) and avoids relying solely on individual technician expertise. Benefits Cognitive ITSM delivers both quantitative and qualitative improvements: Greater operational efficiency through elimination of manual tasks. Faster, more accurate diagnostics. Advanced visibility into service status via intelligent analytics. Scalability without increasing headcount. Smoother, frictionless employee experience. Proactive prevention instead of reactive resolution. Technical challenges and considerations Deploying AI in ITSM comes with specific technical challenges that Telefónica Tech addresses as part of the project. These are not obstacles, they are critical design elements for successful adoption. Implementing AI in ITSM is neither automatic nor trivial. It requires a strong technical foundation and a rigorous approach. One of the most critical factors is CMDB (Configuration Management Database) quality. If asset, relationship, and dependency data is not properly structured, models cannot correlate incidents with real business impact. Another key aspect is model lifecycle governance. Algorithms must be monitored, updated, and audited regularly to ensure reliability, traceability, and alignment with operational objectives. Automations must be designed with clear security, error control, and rollback criteria. Automating without control is not efficiency, it’s risk. Lastly, coexistence between humans and autonomous agents requires a clear definition of responsibilities. AI does not replace human teams. It amplifies their decision-making capacity when properly integrated, optimizing processes and ensuring effective, coordinated incident and service management as part of our Telefónica Tech model. These are not obstacles, they are essential design considerations for sustainable adoption. Conclusion: towards business-aligned cognitive ITSM AI is redefining how we understand IT service management. It’s no longer just about resolving tickets, it’s about anticipating problems, automating decisions, and transforming IT into a strategic business enabler. At Telefónica Tech, we put AI at the service of ITSM by integrating technologies like BMC Helix ITSM within a robust model of intelligent automation, advanced support, and continuous evolution. The result is a more efficient, secure operation, ready for future challenges. Cloud Intelligent Workplace From contact center to intelligent service: 10 key insights into its transformation September 8, 2025