An Agent-Based Approach to the Automation of Risk Management in IT Projects

The failure of projects in the field of Information Technology (IT) is a common problem. The aim of this work was to analyse and suggest a programmatic solution to improving the success of project outcomes. This was achieved through analysis of project variables to identify and quantitatively evaluate the risks surrounding a project and suggest possible mitigations to the project manager. The solution makes use of agent-based systems to monitor project risks and recommend mitigations in response to changing project variables. A knowledge base of common risks and related mitigations was collected through research and a selection of industry professionals in the field of IT with a wide range of experience in various areas of IT. This was used to provide a set of real world test data of risk/mitigation mappings to allow the suggestion and ranking of project risks based on a set of quantitative measures of risk likelihood. An impact for the risk is set by the project manager which, when combined with the calculated likelihood, can allow the calculation of a risk priority. The solution implementation takes two forms; a management tool for the project manager and a monitoring tool, which makes the project statuses visible to all in the project to aid in communication among the project team and stakeholder management, which are two common factors of failure. The research results were positive and showed that the use of real world data in combination with an identified set of measures can identify risks and propose mitigations as well as facilitating communication improvement among project team members through the use of the website monitoring system. This encourages a clear and open approach to project, stakeholder and team management thereby increasing the likelihood of project success.