Data Space and LLM Enabled Decision-Making Support System: An Application in Drug Development

In this study, we analyzed the demand and pain points of multi-business pipelines integrated decision-making scenarios (such as drug development), including the subjectivity, non-interpretability and low efficiency of traditional expert experience-driven decision-making, and the inability to handle a large amount of data. Then, we designed a framework for the Data Space and Large Language Model (LLM) enabled decision-making support system. By integrating Data Space, the framework helps enterprises to achieve multilateral data-driven evaluation and analysis; By integrating LLM and Ai-Agent, collaborative decisionmaking of multi-business pipelines can be realized. This framework can provide quantitative analysis for experts and improve the accuracy and efficiency of decision-making. The effectiveness of this study has been verified in the decisionmaking scenario of drug development. The follow-up studies will further explore the autonomous decision-making of multi-business pipelines and the intelligent conflict resolution mechanism to reduce the dependence on expert experience.