Artificial Intelligence Supporting DevOps Practices in Enterprises: A State-of-the-Art Analysis on Enterprise Clients and Improvement Proposals

The rise of Large Language Models (LLMs) and Artificial Intelligence (AI) is transforming and improving software development, introducing new ways of making and maintaining software by enforcing human-machine collaboration. This thesis explores the current adoption of DevOps methodologies in companies, with a particular focus on the banking and finance sectors. Since customers have widely adopted technologies such as smartphones and the internet, the ability to deliver high-quality, secure and fast digital services has become essential for enterprises to remain competitive. The primary goal of this research is to investigate how DevOps is currently implemented within the banking sector, identify key challenges and limitations, while also proposing innovative solutions that integrate AI to improve development workflows. The study is structured into three main parts. First, it provides a depth analysis of DevOps practices within the financial sector by studying three enterprise clients: a major Italian bank, a large European bank and an Italian banking group. These cases analysis are used to identify obstacles and difficulties that enterprises must face including strict security requirements, regulatory compliance and organizational resistance to change. Second, using the information collected, the thesis proposes and shows AI solutions to automate repetitive tasks, enhance code quality and speed up software release processes. Different AI tools are examined for their ability to support developers in tasks such as code generation, debugging and compliance validation. Third, a practical application is developed to demonstrate how AI can be used in the DevOps life-cycle. The tool assists developers by providing intelligent suggestions, automating compliance with standard laws and increasing overall efficiency. The results are analyzed to assess the impact of AI integration on development productivity and process consistency. The findings show not only the current state and challenges of DevOps in industries but also the potential of AI to help develop and release cycles. Finally, the thesis suggests future development paths, with a focus on continuous improvement and increased integration of AI in DevOps.