We present a proof of concept for utilizing conversational assistance to streamline observability tasks in Kubernetes clusters. Our approach leverages fine-tuned Large Language Models (LLMs), enabling users to interact conversationally with an intelligent system. This system acts as an assistant, supporting users in monitoring, debugging, and troubleshooting Kubernetes clusters and applications.

The intelligent system provides monitoring capabilities, empowering users to track essential metrics and performance indicators like monitoring of resource utilization, pod health, and other significant metrics, enabling users to proactively detect and address issues. Additionally, the system assists users in debugging K8s clusters, offering guidance and recommendations for identifying and resolving issues efficiently.

This approach simplifies monitoring, debugging, and troubleshooting processes, empowering DevOps teams to enhance cluster performance and streamline observability tasks.