The integration of generative AI into financial advisory systems faces significant challenges, particularly in delivering real-time, contextually accurate, and ethically sound investment advice while addressing issues like hallucinations. This paper introduces an agentic Retrieval-augmented Generation (RAG) artefact designed to enhance equity investment decision-making through a multi-agent architecture. Employing a directed acyclic graph (DAG) framework, the system integrates dynamic financial metrics, such as the Piotroski F-Score, to enable robust, scalable, and domain-specific data retrieval and analysis. Using the Design Science Research Methodology (DSRM), the artefact addresses the limitations of traditional advisory systems, such as static models and limited adaptability. This paper aims to contribute to the theoretical understanding of agentic systems in Information Systems (IS) while providing practical insights into their application in financial contexts. The findings highlight the transformative potential of RAG in high-stakes domains and establish a foundation for further exploration of agentic architectures across similar settings.