SUPERWISE for healthcare
Use AI in patient care without losing control of patient data
Why it matters
Accuracy alone is not enough
Patient-facing, operational and clinical AI needs explainability, audit trails, policy enforcement, privacy controls and oversight before anyone relies on it. Read why AI in healthcare fails, and how to fix it.
Clinicians need answers they can check
Reviewers will ask for the record
How SUPERWISE helps
Keep patient work on a model you control
Two controls for patient work, and guardrails on the AI your teams build.
A hosted model for patient work
Sentinel for the outside tools staff use
Guardrails on the AI you build
Proof
Governed AI already supports care teams
Renova Health: AI that equips chronic care managers
Renova Health’s Chronic Care Managers needed evidence-based recommendations without losing the human connection patients depend on, or exposing protected health information.
AI assistants now equip care managers with the right information at the right time, and every AI decision is logged and traceable to its source.
“It’s not about replacing our CCMs, it’s about equipping them with the right information at the right time.”
In medical imaging, NuevoSono embedded governed computer vision in its next-generation IVUS platform, increasing interpretive consistency across operators. Read the imaging story.
The details
What your compliance team will ask
The hosted model, the BAA, what Sentinel replaces, the guardrails, and both customer stories in full.
The hosted model
- Patient work goes to a model SUPERWISE hosts on its own equipment, under a BAA with your organization.
- Hosted by SUPERWISE in a secure cage at 365 Data Centers, Nashville, Tennessee.
- The model is gpt-oss-120b. Subject to change as models advance.
- Staff reach it through SUPERWISE Chat, or through Sentinel configured to route a tool to it.
It is available now under a custom contract. Ask about a custom contract.
The BAA: who signs and what it covers
SUPERWISE signs a Business Associate Agreement with healthcare customers, and the scope is confirmed during scoping. The security and BAA facts are all in the Trust center.
- Any covered entity can sign, including practices, hospitals, and health systems.
- Covered: Chat on the hosted model under a custom contract; Sentinel routing to the hosted model.
- Not covered: Self-serve Chat plans; The platform (AMP); Outside models a customer chooses; Web search.
- If we confirm a security incident that affects your data, we notify you within 72 hours.
- The BAA carries the same 72-hour notice.
Self-serve SUPERWISE Chat plans don’t include a BAA. Keep PHI out of them.
Want the agreement in hand first? We sign a BAA with healthcare customers; ask for our template.
What Sentinel replaces
SUPERWISE® Sentinel inspects supported prompts on the PC and replaces the identifiers it catches before the request leaves. ChatGPT or Claude receives the transformed prompt without those replaced values. Live activity shows which rule fired.
Sentinel replaces email addresses, phone numbers, Social Security numbers in the 123-45-6789 format, credit card numbers, IBANs, IP and MAC addresses, full-length OpenAI keys, and AWS secret keys when labeled. Bank account numbers, tax IDs and EINs, national IDs, VINs, dates of birth, and medical record numbers are replaced only in the labeled formats the gateway recognizes. The admin chooses whether Sentinel redacts or blocks, and sets the replacement text. For each request Sentinel records metadata only: counts, which rule fired, and latency. Prompts go straight to the AI provider; SUPERWISE receives only check metadata.
Each organization determines its own HIPAA compliance, and each tool’s HIPAA status comes from its own agreements. Sentinel is one safeguard inside that program. Read the answers for ChatGPT and Claude, and check the PII redaction documentation before relying on a format.
Guardrails and the audit log
SUPERWISE applies configured guardrails, records checks, and supports policy controls.
- AMP guardrails: The AMP platform applies configured PHI rules before model calls. The platform (AMP) is not covered by a BAA. Keep PHI out of it.
- Audit log: Each check is recorded as metadata: when it ran, which guardrail fired, and the outcome. Prompt content stays private.
- Listed identifiers replaced: Sentinel replaces listed identifiers before a prompt reaches a public AI tool.
- Runtime guardrails: Policies are enforced at runtime and recorded after deployment.
Healthcare-specific guardrails
- Redaction of labeled medical record numbers and dates of birth
- Medical misinformation filtering
- Clinical safety boundaries
- Audit trails for regulatory review
- Explainability requirements
Operational AI governance in healthcare systems (Renova)
The SUPERWISE AMP platform provides the governance layer for operational AI in chronic care management, enabling real-time patient data integration, evidence-based recommendations, and configured PHI guardrails before model calls.
The platform (AMP) is not covered by a BAA. Keep PHI out of it.
Every AI decision on the governed workflow is logged.
Governance and oversight
- Oversight across data pipelines and decision-support workflows with integrated EHR connectivity
- Explainability for AI-assisted care recommendations, enabling clinicians to understand model reasoning
- Audit trails for every AI decision, providing complete documentation for regulatory compliance
- Risk controls aligned with clinical and operational guardrails, preventing PHI exposure and ensuring patient safety
- Real-time monitoring and observability across live healthcare operations
Governed AI for high-resolution intravascular imaging
Clinical challenge
- Coronary artery disease assessment depends on precise plaque visualization and lesion assessment
- Interpretation variability across operators impacts treatment decisions
- Regulated clinical environments require oversight and compliance for AI-assisted diagnostics
AI and governance approach
- Computer vision models embedded in high-resolution intravascular ultrasound workflows
- Explainability and observability built into deployment architecture
- Guardrails aligned with clinical and regulatory expectations for security and compliance
Operational impact
- Improved interpretive consistency across clinicians, reducing variability in plaque detection and lesion assessment
- Streamlined cath lab workflows with decision support aligned to clinical outcomes
Real-world validation
- Deployed with NuevoSono’s next-generation IVUS platform for coronary plaque detection and lesion analysis
- Increased consistency in image interpretation across operators
- Streamlined clinical workflows with decision support aligned to outcomes
- First-in-human studies initiated in 2025, advancing toward clinical commercialization
What SUPERWISE provides
- Explainability for model decisions and image analysis
- Observability across deployment pipelines and workflows
- Guardrails aligned with clinical and regulatory expectations
- Audit trails for regulatory review and compliance
- Risk controls for security and clinical reliability
See it in action
Watch guardrails catch PII in a symptom checker
Both demos use a symptom checker built on the SUPERWISE platform.
Symptom checker with PII protection
Watch a guarded symptom checker built in minutes, and see how SUPERWISE guardrails check its requests. Watch it on its own page.
Preventing data leaks
Watch SUPERWISE guardrails catch PII in a symptom checker built on the platform. Watch it on its own page.
What our experts build: SUPERWISE Council
A healthcare team describes the AI system it needs in plain English. Our experts turn that brief into a governed system, with its review record, guardrails, versions, monitoring, and traceability built alongside it and checked against NIST AI RMF, HIPAA, FERPA, and FINRA controls.
Talk to us
Talk to our team
Tell us what patient work you want AI to help with. The services team will suggest the right first step.