Aplikace nástrojů generativních AI aplikací pro finanční plánování

This thesis analyzes the current capabilities of generative artificial intelligence tools, specifically large language models developed by OpenAI, within the context of corporate financial planning. The thesis evaluates not only the accuracy of the generated outputs but also their relevance to the company and their potential for future real-world implementation. The practical section of this thesis consists of case studies focusing on two companies within the non-alcoholic beverage industry. The artificial intelligence was tasked with developing relevant and up-to-date financial plans for these companies based on corporate and economic data. The plans were generated using three distinct variations of the same AI tool ChatGPT 5.4, the base model, a trained model, and an agentic model. The results indicate that the formulated recommendations and theories lack deeper analytical insight, and the outputs remain inconsistent across various iterations. However, the research also suggests that with additional calibration of the agentic model, it was possible to achieve a higher degree of relevance compared to both the base and trained models.