In order for an agent to be considered both an interface agent and autonomous, it follows that there must be some part of the interface that the agent must operate in an autonomous fashion. The user must be able to directly observe autonomous actions of the agent and the agent must be able to observe actions taken autonomously by the user in the interface The ability of a software agent to operate the same interface operated by the human user, and the ability of a software agent to act independently of, and concurrently with, the human user will become increasingly important characteristics of human-computer interaction. Agents will observe what human users do when they interact with interfaces, and provide assistance by manipulating the interface themselves, while the user is thinking or performing other operations. Increasingly, applications will be designed to be operated simultaneously by users and their agents [1]. This study is motivated by the need to solve a problem of human resource optimization in the first year informatics practical course as presented by the R.A.U. Standard Bank Academy for Information Technology. The major aim being the development of a prototype system capable of automatically grading first year Microsoft Visual Basic.Net applications. The prototype system will ultimately render assistants obsolete in the grading process and free the assistants to help students with problems related to the informatics course. Developing the envisaged prototype requires much preliminary reading on artificial intelligence and its applications, more specifically autonomous interface agent architecture. Case-based reasoning and machine learning has been identified as having great potential and applicability in the development and implementation of the envisaged prototype and for this reason these topics will provide a foundation on which to build this dissertation. / Ehlers, E.M., Prof.