This paper advances the hypothesis that a business can be regarded as a collection of interacting rule based processes, that are analogous to a set of rule based coordinating pervasive computing agents that make up intelligent environments. In this paper we explore this hypothesis and, in particular, investigate the application of directed graph theory, normally used in determining the stability of systems of pervasive computing agents, to business systems. In doing this we present an account of interaction networks, business process reengineering, show how they are both based around the use of 'process' abstractions and illustrate how they can be integrated using a representative example. This paper represents a first step in our longer-term goals to explore the full potential for AI to create smart business monitoring and management tools.