LLM-Driven Continuous Workflow Personalization in Multi-Stakeholder Environments

Recent applied work (e.g., DIP AI-Driven Project Management and LLMs in breast cancer management) demonstrates the value of LLMs in automating and optimizing specific tasks. However, workflow personalization—particularly in multi-stakeholder environments—remains underexplored. This research would investigate how LLMs can continuously “learn” from workflow interactions, adapting both the information presented and the steps taken to fit the evolving needs, expertise, and preferences of various users. For example, an LLM could provide technical logs and data quality metrics to engineers, summary analytics to managers, and regulatory alerts to compliance officers—all within the same workflow but personalized in real-time. This goes beyond static role-based access to truly adaptive, context-aware workflow experiences, increasing both efficiency and user satisfaction.

References:

1. DIP AI-Driven Architecture for Enhanced Project Management Using Large Language Models. Roseline Florence Gomes, Lijo Thomas (2024). 2024 IEEE 4th International Conference on ICT in Business Industry & Government (ICTBIG).
2. Utilizing large language models in breast cancer management: systematic review. Vera Sorin, Benjamin S. Glicksberg, Y. Artsi, Y. Barash, Eli Konen, Girish N. Nadkarni, Eyal Klang (2024). Journal of Cancer Research and Clinical Oncology.