• Support teams lose time through manual summaries, recurring answers and context switching
• Knowledge articles and ticket data are not always consistent enough to reliably support AI
• Without clear modules, governance and enablement, AI remains a pilot rather than a stable part of the workflow
• Modular implementation: current-state analysis, configuration in Service Operations Workspace, workflow adjustments (summarisation, AI answer generation, knowledge suggestions), coaching and iterative fine-tuning
• Use-case selection and prioritisation for quick wins in support (e.g. ticket summarisation, answer suggestions, knowledge support)
• Guidelines and governance for quality and usage (e.g. review logic, responsibilities, feedback loops)