Prompt Patterns for Agile Software Project Managers: First Results

Prompt Patterns for Agile Software Project Managers: First Results

연구 분야: Software Development

학회: International Conference on Software Business

In the evolving field of Agile Project Management (APM), the role of the project manager is in transition. This paper identifies common ‘pain points’ in APM through a literature review and constructs a theoretical model to address them. The study introduces ‘Prompt Engineering’ as a novel approach to leverage artificial intelligence (AI), specifically ChatGPT, for mitigating these challenges. Empirical research evaluates ChatGPT's capabilities and reliability in managing various project tasks using engineered prompts. The findings suggest that while ChatGPT cannot fully replace human project managers, it excels in assisting, guiding, and automating specific tasks when guided by well-crafted prompts. As an outcome, prompt engineering patterns for project managers is proposed to facilitate the application of AI in agile settings. In this paper, we introduce patterns for requirements management, stakeholder and management teams and role clarification. The paper concludes that ChatGPT's knowledge is generally reliable but emphasizes the need for expert evaluation in critical areas.