

See Why GenAI Workloads Are Breaking Observability with Wayne Segar
13 snips Jun 26, 2025
Wayne Segar, Director of Global Field CTOs at Dynatrace, dives into the complexities of monitoring unpredictable AI workloads. He discusses the breakdown of traditional observability practices when applied to non-deterministic AI systems. The conversation highlights the importance of ‘human in the loop’ approaches and how AI Centers of Excellence help navigate compliance challenges. Wayne also touches on the benefits of consolidating observability tools and the innovative drive of startups compared to larger enterprises.
AI Snips
Chapters
Transcript
Episode notes
Complexity of Observing GenAI Workloads
- GenAI workloads are non-deterministic, making their outcomes and performance unpredictable.
- Monitoring must now include cost and performance metrics of AI models, increasing complexity significantly.
AI Chatbots Fail Customer Service
- The user experience of AI chatbots often frustrates customers due to lack of human response.
- Even simple responses from robots like chastising repeat emails deteriorate customer satisfaction.
Keep Humans in AI Loops
- Maintain a human in the loop for AI-based operations to avoid full automation pitfalls.
- Use AI to suggest solutions while letting humans make the final changes to ensure control and quality.