This chapter demonstrates how to use Google’s Gemini model to design and implement a Clinical Decision Support System (CDSS) driven by GenAI. The system allows clinical workflows to use multimodal reasoning that considers the context by combining text and image data, such as imaging studies and patient reports. Our framework uses Gemini’s advanced language and vision capabilities to understand both structured and unstructured input, route queries through LangGraph logic, and call on domain-specific tools to give medically relevant answers without crossing ethical lines. The case study shows how GenAI improves the accuracy of decisions in areas like symptom triage, patient education, and workflow navigation, all while following strict healthcare safety rules. Implementation of key features include graph-based state transition, system prompt management, and measurement metrics to establish model performance in safety, tone, and medical accuracy. We also define challenges to applying generative models in clinical practice, e.g., handling ambiguity of patient input and aligning AI output with provider accountability. This ride provides grounds for bringing GenAI into intelligent healthcare systems—driving conversational agents from scripted responses to service-based adaptive decision-making.