The Decision Network Analyst

When AI agents replace human personnel and make independent decisions, their utilisation becomes not only a technical issue but also a governance issue. By making decisions on behalf of an organisation, AI agents have delegated authority. A central point of The Decision Network Analyst is that a system cannot be held accountable for a poor decision made by an AI agent. Ultimately a person or group will be held accountable, and this is potentially an organization’s board of directors. This book is for the people who must act on that point: the business analysts, risk and governance professionals and transformation leads, who are asked to put agents into workflows and find that the usual process-mapping toolkit stops at the decision itself. Directors and executives will find the book useful for a different reason: it shows what good analysis of a decision point looks like, so they can ask for it. It defines a role of the Decision Analyst, whose work is to examine a decision point before an AI agent is deployed in it, and a second role, the Decision Network Architect, who oversees the network of agents across the organisation and keep the record of what has been delegated to them. Guiding how to deploy AI agents responsibly in critical decision points, the book helps analysts scrutinize AI agent configurations prior to implementation, thereby reducing potential risks and cost overruns. It also discusses how network architects can advise management on the strategic deployment of agent networks so that the agents best achieve a firm’s goals and objectives.