Evolution of the Finance Department through Artificial Intelligence: Digital Assistants and organizational impacts

Until 2020 the applications of artificial intelligence remained confined to the sphere of scientific experiments or concentrated in vertical areas, which could not be governed directly by the final user. Over the last few months it has been made available to a much wider public, both in the consumer and business applications, with the possibility of direct interaction by users in the form of chat. Some analysts define this historical moment as a new industrial revolution, which will change the way we live and work, without however clarifying specifically what the positive and negative consequences will be, especially in the area of business impact. It will probably be a disruptive process in some areas and this thesis project aims to explore the potential that artificial intelligence may have in the field of business support, analysing advantages and possible organisational impacts, with a specific focus on the world of Finance. One of the latest innovations in the field of artificial intelligence is Digital Assistants. An Artificial Intelligence Assistant is a software system designed to simulate human interaction and assist people with various tasks. These assistants, often based on machine learning algorithms and powered by natural language processing (NLP), can be used in a wide range of contexts, including customer support, information management, business process automation and data analysis. Like all artificial intelligence algorithms, they have all knowledge based on public information, but unlike public interfaces, they are embedded in a private environment, to which only certain users can have access. This makes it possible to proceed to a specific training process, to integrate some private information and obtain a real Private AI, safeguarding the confidentiality of some information that do not want to make public. The training process for a Company Owned Assistant is based on the customisation of a pre trained artificial intelligence model, to respond more specifically and relevantly to the needs and preferences of the user and the company. This phase also includes the possible development of specific integration interfaces with other corporate information tools, with which the assistant may need to relate (e.g. ERP, CRM, BI, EPM, MES, WMS, DWH, etc.). The digital transformation potential of the AI Assistant was tested with particular reference to the Finance area, through the development of an application to a real business case. The choice favoured a field in which artificial intelligence can autonomously generate business analysis models, following on demand requests formulated in natural language, with the aim of bridging a gap still present in all business contexts, even the most structured, concerning the inefficiency of traditional decision making processes in contexts of high uncertainty. The introduction of Artificial Intelligence in companies, as in the application case developed, could also deeply transform the organisational structure and roles within them. This change may impact several organisational levels, requiring a review of work processes, employee interactions and responsibilities. This study explores the consequences of the integration of Digital Assistants in the corporate context, focusing on four main areas: organisational impacts, possible roles at risk, the introduction of possible new figures and ethical aspects.