Data Intelligence as a Service: A Cloud Platform for AI-powered Cybersecurity Research and Practice

This research project is a collaborative work among researchers and students at East Carolina university and McMaster University in Canada to establish Data Intelligence as a Service (DIaaS) cloud platform for AI-driven cybersecurity research, development, and hands-on practice to train security analyst from different industries. The DIaaS utilizes Google Cloud Platform (GCP) to provide customized virtual workstations for the security analysts based on their needs to practice and promote their knowledge of cloud and AI-based cybersecurity solutions on synthetic or proprietary dataset. The platform will generate realistic behavior-pattern driven datasets representing normal and RED-team activities that will be augmented significantly via our proprietary generative adversarial network solution (GAN). The locally developed AI solutions will be packaged into docker images which will be offered through the virtual workstations. The vulnerability analysis tools will leverage explainable artificial intelligence (XAI), inverse reinforcement learning, Transformers, knowledge-graph visualization techniques, as well as cloud-based process automation using CI/CD and Kubernetes container orchestration. Such solutions will allow the security analysts to develop their own ML models, migrate to cloud, and easily interact with the ICS using dialogue-based chatbots.