CalmWave
CalmWave combines bedside monitor signals, medications and EHR activity to help hospital teams manage alarm burden and understand patient changes. Its machine learning system proposes alarm thresholds with supporting data, while nurses retain control of the bedside monitor and can reject recommendations. Clinical safety gates and physician orders constrain when an adjustment can be suggested.
Operational views extend to device performance and hospital activity. Wellstar Kennestone's CCU provides a named deployment for the alarm workflow. The newer Tessera service offers early access to queries across deidentified clinical data for research. The two offerings serve different tasks: bedside operational support and aggregate data exploration.
Capability Axes
An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read
Machine learning drives alarm optimization and patient specific recommendations. The platform also combines monitor, medication and EHR data for operational analysis, giving hospitals a shared view beyond the recommendation engine.
Recommendations require stable vitals, no recent intervention, no threshold change for three hours and repeated alarms without instability. Nurses review the rationale, adjust monitors manually, or reject with a reason. Physician orders bound permitted changes.
Transparent AI uses machine learning with structured vital signs, medication, staffing and EHR activity. Recommendations expose their rationale and apply clinical safety gates. Model identifiers, training dataset composition and version change history remain unspecified.
CalmWave runs its own Transparent AI algorithms on Oracle infrastructure. That identifies the core model development path. Tessera's language model layer and the complete provider inventory, retention terms and training rights remain unspecified.
Wellstar Kennestone's CCU account describes a six week rollout and a reduction from 68 to 36 alarms per hour. The account lacks a dated measurement period, patient denominator and controlled comparison, limiting interpretation of the reported benefit.
Training uses deidentified protected information, with identity restored only within hospital environments. Recommendation and incident logs support review. Model retention deadlines and a detailed AI safety event response process remain unspecified.
HIPAA controls accompany a marketplace service agreement with data protection terms, restricted data use and incident notification duties. A readable BAA with its applicable tier and obligations for PHI subcontractors is not established.
The trust center identifies SOC 2 Type II and lists the report behind an access request. Annual independent auditing is asserted. The audit artifact, covered period and detailed scope remain unverified.
The product supports alarm configuration and hospital operations, leaving bedside changes to clinicians. Its stated scope excludes curing, mitigating, treating or preventing disease. This positioning does not establish clearance for autonomous diagnosis or treatment.
Medical, nursing and ICU leaders review models before deployment. Quarterly audits address accuracy, bias and drift, with technical and clinical review. Comparative subgroup results and an independent model behavior audit remain unspecified.
The marketplace service agreement provides correction and refund remedies for material product nonconformance. These create a vendor remedy route, but do not expressly guarantee individual alarm recommendations or compensate for clinical consequences.
Wellstar's deployment combines live Philips IntelliBridge and Epic data, including medication administration and vital signs. Nurses change bedside monitor settings themselves. The named connection establishes production data intake without demonstrating bidirectional EHR writing.
The service runs on Oracle Cloud Infrastructure following a migration from its earlier hosting arrangement. A marketplace subscription remains available. Committed processing regions, tenant isolation and cross border data handling are unspecified.
The marketplace contract costs $7,152 per ICU bed for 12 months and includes a purchase route. This establishes the standard alarm management subscription; Tessera's early access data service has separate, unspecified commercial terms.
Critical care deployment at Wellstar Kennestone supports the alarm management use case. Broader positioning covers patient flow, discharge readiness and hospital operations, with less validation for those uses than for the CCU workflow.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
| Entry Price | Pricing Basis | BAA Tier | Implementation | Source |
|---|---|---|---|---|
|
$7,152 per ICU bed per year
$7,152 baseline
|
Annual subscription per ICU bed through AWS Marketplace. | HIPAA controls claimed; readable BAA and tier unspecified. | Implementation and onsite support described; separate fee unspecified. | Vendor Published |
The standard contract runs for 12 months. Tessera early access pricing is unspecified.