Recently, I was involved in automating the preparation of IT incident reports by integrating monitoring solutions and generative AI. This post is to share my observations during this ongoing exercise. It outlines how to leverage the power of the AI service to interpret alerts generated by an observability solution (in this case, eG Enterprise Suite).
It utilized the Gemma AI model, running locally via Ollama for on-premise users, and the Llama-3.1 AI model with OpenAI for those with internet connectivity. The outcome of this exercise resulted in faster understanding of system issues and potential remediation steps. It automates the incident report creation by a Level 1 engineer.