As generative AI continues to reshape the technology landscape, its impact on observability is becoming increasingly profound. In this talk, Diana Todea examines how AI agents can extend traditional observability workflows and how observability, in turn, can be used to evaluate and optimize the performance of generative AI systems.
Todea begins by exploring the architecture of transformers and generative models, showing how these foundational components integrate to create AI agent workflows. She then highlights key use cases in which generative AI enables fully automated operations, improving response times and reducing manual effort. The session also introduces the emerging concept of using large language models (LLMs) as “judges” to make contextual decisions within observability frameworks. Attendees will leave with a deeper understanding of how generative AI can be practically applied to DevOps platforms and how machine learning principles can elevate the observability and reliability of modern systems.