
Gemini for Healthcare: A Governed Multi-Step Clinical Workflow Platform

Google Cloud announced Gemini for healthcare, a purpose-built industry solution extending the Gemini agent introduced last month at Gemini at Work. The stated problem: healthcare runs on fragmented data across disconnected EHRs, claims systems, and administrative tools, and clinicians spend nearly 28 hours a week manually reconciling information. The argument is that a general AI model alone is not enough; healthcare needs a secure, governed orchestration layer that coordinates work across existing systems, assembles verifiable evidence, and runs multi-step workflows with human oversight.
The solution has four components. Skills are reusable instructions written by domain experts that guide the agent through tasks such as medical records synthesis, pre-appointment preparation, coverage and policy review, and payer benefit determination; teams can build custom skills for their own policies and standard operating procedures. Connectors link Gemini to EHR, claims, care management, and administration systems while preserving existing identity and role-based access controls, so data does not need to be migrated and permissions are not altered. An open partner ecosystem includes systems integrators and digital health partners. The platform layer includes centralized policy management, audit logging, verifiable citations, and support for HIPAA compliance, and it supports open model choice: frontier models for complex clinical reasoning and fast, cost-efficient models for routine administrative steps under the same governance. Every agent operates with a distinct identity, inherits the user’s permissions, and enforces need-to-know boundaries across users, sub-agents, and workflows.
The announced workflows focus on two points. At the point of care, Gemini synthesizes a longitudinal timeline from EHR and external records, separates active problems from resolved ones, lists allergies, flags medication discrepancies, prepares charts before scheduled visits, and assembles authorization packets aligned with payer policy before submission, flagging documentation gaps and drafting appeals when denials occur and the record supports it. At the point of coverage, it checks requests against the plan’s medical policies, organizes findings for clinical reviewers, parses benefit booklets to estimate out-of-pocket costs, track deductibles, and verify in-network status, and applies Medicare coverage criteria for Medicare Advantage members when a plan’s internal rule is stricter.
The connector list covers EHR and clinical data (Epic EHR and Oracle Health with read-only access and no write-back, InterSystems FHIR MCP Server, Verato for identity resolution, Keragon for 400+ healthcare applications), eligibility and claims (Optum Coordination of Benefits and Optum Network Eligibility in limited release, TriZetto Provider Solutions, GoodRx, Infinitus Relay), terminology and coding (IMO Health, ICD-10 Codes, RxNorm Drug Vocabulary), coverage and provider reference (CMS Coverage, NPPES NPI Registry), evidence-based guidelines (Optum EvidenceVue, BMJ Digital Health), and clinical/operational systems (Salesforce Health Cloud, Suki, UiPath, Synthpop Careflow, Alteryx, Sana from Workday, ServiceNow, Microsoft 365, Google Workspace, Cyera Dataport). Implementation partners listed include Accenture, Deloitte, PwC, and many others.
Early collaborators mentioned include HCA Healthcare, Northwell Health, Optum Insight, Blue Cross Blue Shield of Massachusetts, and ASCO. Google Cloud says clinical data, prompts, and generated outputs remain within the organization’s governed environment and are never used to train or fine-tune Google’s foundation models. Gemini for healthcare is available in preview in the U.S. starting today.

