Best Practices Implementing AIOps in Large Organizations

This paper is a Qualitative Research Narrative Case study in which the researchers were total participants, and federal contractors, focused on an essential but underrepresented area, the Artificial Intelligence for IT Operations (AIOps) space, including the need for understanding interdependencies, technology adoption curve, and consolidation of technical silos. The research question is, "Do the observations and recommendations surrounding AIOps from commercial enterprise literature and use cases apply to AIOps executed in the government space by federal contractors?" This paper will focus on first a literature review from commercial sources on AIOps, understanding the interdependencies and what technology leaders need to go through for the adoption curve, and then examine merging data science, IT operations, and software development. Then, the paper will provide Veterans Engineering's (VE) related work experience and our observations. Finally, in this paper, we will discuss future work, our recommendations for best practices, and a conclusion to the initial question we asked in this research paper. As researchers and practitioners in the AIOps space, Veterans Engineering (VE) works with federal agencies. It has provided observations and lessons learned from our recent work and research to foster best practices in the AIOPs space. VE's top observations appropriately estimate the interdependencies and the resulting level of effort, including competencies in soft skills, process engineering, political capital, communication, and authority to make the initiative a company or agency-wide effort. VE recommends focusing on the skill domains rather than the tool selection. AIOps connects business processes and values to automated actions; focusing on business objectives first and tool selection second is critical to overall program success. Both the federal government and private sectors face similar challenges and success factors when implementing AIOp tools, such as the need for skilled engineers, dealing with large-scale data, and interoperability issues among tools. Success often hinges on having a senior leader champion the effort and a dedicated team. The federal government could learn from successful private sector practices to navigate these challenges more effectively.