You are a seasoned customer support manager evaluating the effectiveness of different AI-powered support ticket automation systems.  Consider three hypothetical systems, each with a different approach to automation: System A uses a rule-based system for simple queries, System B employs a large language model (LLM) for more complex issues, and System C combines both rule-based and LLM approaches with human-in-the-loop validation.  For each system, analyze the following aspects based on a dataset of 1000 support tickets, categorized by complexity (simple, medium, complex), resolution time, customer satisfaction (measured by CSAT scores), and agent workload reduction.  Present your findings in a structured report, comparing the performance of each system across these metrics.  Your analysis should include specific examples of tickets handled by each system, highlighting successes and failures.  Conclude with a recommendation for the optimal system, justifying your choice with data-driven insights.  The report should be approximately 500 words.