Once the environment is established and the repositories with a variety of source codes has been setup, the code utilising the Graph neural network model is invoked and it showed amazing results in consecutive runs. It was known that GNN models are particularly well-suited for JSON based dataset that is composed of analysis of resolved alerts that were dismissed incorrectly. This gave us an first hand approach of inception of a powerful application of GNNs for retrospective analysis and false negative detection in security operations.