Design and Development of an AMS Ticketing Portal with AI-Based Summarization for Knowledge Reuse

In enterprise service management, customer communication and incident tracking often rely on fragmented channels, making it difficult to reconstruct issue histories and reuse past solutions. Currently, the company does not have a dedicated tool to centralize ticket management and the related communication with customers. As a result, information was scattered across long email threads and it was often difficult to quickly understand the history of an issue or to reuse solutions from past cases. To address these limitations, this thesis presents a web portal developed to enable customers to authenticate, submit new tickets, browse and filter existing requests, and access detailed ticket pages that include the full interaction history. The front-end of the portal is built with Angular 20, while Microsoft Dynamics 365 CRM is used as the back-end to store and manage tickets, users, and email exchanges. The most distinctive feature of the portal is the integration of an AI-based summarization function directly into the ticket detail page. For each ticket, the user can request a summary of the conversation with a single action. This request triggers a Power Automate flow connected to Dynamics 365, which collects all emails and descriptions linked to the ticket, processes their content, and calls an AI model to generate a short recap of the discussion. The resulting summary is then returned to the portal and presented as a compact overview of the reported issue, the actions taken, and the current status. In this way, operators no longer need to read through entire email threads to understand a case. At the same time, the summaries themselves become structured knowledge items that can be searched and reused when dealing with similar tickets in the future. The thesis presents the main architectural choices, the requirements and the implementation details of both the portal and the AI flow. It also discusses the qualitative benefits observed in terms of easier access to information and better reuse of existing knowledge: the summaries generated for each ticket gradually form a lightweight, structured knowledge base that captures not only the final solution, but also the context of the problem and the intermediate steps taken to resolve it. This makes it easier for operators to search past cases, identify similar issues, and apply previously adopted solutions, while also supporting onboarding of new team members and promoting a more consistent approach to incident resolution over time.