Healthcare and life sciences (HCLS) organizations generate large volumes of patient-level data spanning genomic, clinical, medical imaging, and other modalities, but analyzing it at scale remains challenging due to distinct storage infrastructure and domain-specific formats (FHIR for clinical records, DICOM for medical imaging, and VCF for genomic variants). This two-part blog series shows how to build agents that interact with multimodal HCLS data, introducing a data governance layer with automated metadata capture and semantic discovery, and a multi-agent system where specialized AI agents connect to data stores and tools via Model Context Protocol (MCP) servers. The system enables natural language querying and self-service analytics, and extends beyond patient-specific data by adding two external knowledge agents: a clinical trials agent querying ClinicalTrials.gov and a PubMed agent searching biomedical literature. Part 1 creates a framework to ingest multimodal HCLS data into purpose-built AWS services, establishes a unified, governed data catalog using Amazon SageMaker Unified Studio, and creates interactive visualization dashboards with Amazon Quick. Part 2 builds the multi-agent system using Amazon Bedrock AgentCore and Strands Agents SDK and trains predictive models with Amazon SageMaker AI. The use case considers cardiovascular disease risk evaluation, where assessing risk requires synthesizing insights from clinicians, radiologists, and genomics specialists, which typically happens in silos and requires significant manual effort.