10075732Predictive modelling for HEI-Commercialisation DashboardClosedGrant for R&DInnovate UKThe use of technology in everyday life is increasing year on year. One of the methods that is becoming increasingly common to see is the use of Artificial Intelligence (AI) to support or deliver services. Frequently it is used to respond to queries when a human staff member isn't free. Or to process very large amounts of information in an efficient way. However, as it becomes more common additional requirements are needed for products to ensure they are safe for the public and also patients (if used in a healthcare setting). For example, it is known that AI software can have biases built into it without the knowledge of the teams that create them, depending on what information sources are used in their development. This project is designed to support a process known as Artificial Intelligence Assurance. That seeks to ensure AI products produce trustworthy information that is safe, appropriate and reliable while also being in compliance with relevant standards. We propose to build a set of survey systems that can support the auditing, accreditation, certification and impact assessment of AI products. Specifically we will focus on the early stage ventures and spinouts that are produced in Universities. This is because in the UK (and Europe) many companies with a technology component to them (AI in this case) will enrol on a university business accelerator or incubator programme at some point in their commercial journey. This provides a suitable space to be able to test if a company's AI software is safe, reliable and accurate. In addition, the company will have access to budgets and support from the incubator/accelerator programme to fix any issues that may arise during that time. We will build the AI assurance tools (e.g. an audit) into online surveys. This is because it allows us to automate the analysis for each of the assessments and create visualisations of the score i.e. bar-charts, pie-charts etc. Consequently, as soon as the company fills in the paperwork we know what level they are at. This rapidness will be useful to the managers of university incubator/accelerator programmes to keep track of what is happening across the institution, and make any changes that they need. It will also allow them to start to make a university wide plans (roadmap) on what to do with AI-based companies as they can observe trends in dashboard data.