Support teams receive lots of date in form of client logs and traces while working on support tickets. This data can provide very valuable actionable insights if processed correctly. In this collaborative project between IBM and York University, we have created tools to parse client logs in systematic way using LLM. This tool is executed on data arrival and consolidate findings with historic data. Consolidation keeps building pipeline of insights in a prioritize order for development team to act on. As it’s other application, anomalies detected before is passed to another LLM which is trained on specific product domain, provides possible resolution for each anomalies identified. This helps support engineers to head start their PD process saving time to resolution.