Integrating LLM-based AI Agents into Technology Roadmapping: Design Considerations and Opportunities


NAZLIEL K., Kayabay K., KOÇYİĞİT A.

2025 IEEE International Conference on Technology Management, Operations and Decisions, ICTMOD 2025, Glasgow, İngiltere, 20 - 22 Ekim 2025, (Tam Metin Bildiri) identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/ictmod66732.2025.11371870
  • Basıldığı Şehir: Glasgow
  • Basıldığı Ülke: İngiltere
  • Anahtar Kelimeler: Artificial Intelligence (AI) Agents, Large Language Models (LLMs), Technology Roadmapping
  • Orta Doğu Teknik Üniversitesi Adresli: Evet

Özet

The growing capabilities of LLMs and AI agents offer new opportunities for enhancing Technology Roadmapping, a core method in strategic planning. While current applications of LLMs in roadmapping have largely focused on trend identification and content generation, the broader potential of agentic systems across all phases of roadmapping remains underexplored. This paper examines how LLM-based AI agents can support and augment roadmapping activities, and proposes a set of design considerations to guide their effective integration. Drawing on agentic design patterns, technology management principles, and recent empirical studies, we introduce a framework that outlines six core requirements for building usable, adaptable, and explainable roadmapping systems. We also highlight specific opportunities where these agents can be implemented and add value. The findings aim to inform the future development of AI-augmented roadmapping tools and contribute to the advancement of strategic foresight capabilities.