Sovereign by design, enterprise by default
TRIMPACT.
TRIMPACT TRINITY IMPACT
The bedside critical-care platform. TRINITY continuously captures data from the equipment already in the room and scores risk on premises; IMPACT presents it as one clinical view for the unit.
From data noise to actionable insights
TRIMPACT / data flow
From fragmented data to actionable insight
Fragmented signals
- Vitals
- Labs
- Record systems
- Imaging
- Observation
- Medications
- Pose geometry
TRINITY
Captured and scored on premises
IMPACT / clinical view
Early pattern
Brady events clustering: 5 in the last 6 hours
Bed 04. Vitals, monitors, and bedside observation agree.
Why it is flagged
Heart-rate variability falling since 03:00
A pattern associated with late-onset sepsis, shown with its sources.
Into human hands
Surfaced to the bedside nurse for early review
Time-stamped and logged. A clinician reviews; a clinician decides.
Awaiting clinician review
Human in the loop, by design
Every insight is decision support. A clinician reviews it, and a clinician decides.
Sovereign by design, enterprise by default
Capabilities
Three layers, one clinical picture
TRIMPACT is built from three layers that work as one: FIGURA senses the bedspace, TRINITY captures and scores, IMPACT presents and a clinician acts.
Layer 01 · Integration hub
TRINITY
The bedside-grade integration hub that captures every signal in the room and scores risk on the unit's own compute.
A bedside-grade device that speaks to the equipment already in the room. It captures high-frequency physiological data, FIGURA pose geometry, EMR data and clinical context together, then runs models on the unit's own compute.
- InputsMonitors · devices · record systems
- InterfacesHL7 · FHIR · waveform
- ProcessingOn premises only
Layer 02 · Dashboard
IMPACT
The clinical dashboard that turns captured signals into one adaptive, role-gated view of the whole unit.
The visual layer and single source of truth for the unit. Vitals, labs, and AI risk assessments consolidate into a view that adapts to how the team actually works.
- ViewsBed · unit · shift handover
- SignalsRisk scores · trends · alerts
- AccessRole-gated with MFA
Layer 03 · Contact-free sensing
FIGURA
The sensing layer that reads posture, movement, and presence in the bedspace as geometry, never as a face.
Frames are turned into joint and limb coordinates inside the room and then discarded. What reaches TRINITY is structured observation, so continuous positioning and movement data arrive with no imagery to store or secure.
- ReadsPosture · movement · presence
- StoresGeometry only, no imagery
- IdentityNo face, no likeness
Clinical value
Anticipate deterioration instead of reacting to it
- Early-warning insight from multi-modal data rather than single-threshold alarms
- Risk trends that make a patient's direction of travel visible across a shift
- Alert prioritization designed to reduce cognitive load, not add to it
- Validated shadow-mode first, observing and logging only, before any workflow expands
Operational value
Built for the unit and the institution
- Demand and flow views that help charge nurses and coordinators plan capacity
- Quality metrics and audit-ready records for improvement work
- A research-grade local data foundation for institutional studies
- Deployment sized to a single unit first, expanded only on measured results
Deployment and safeguards
01 / On premises
All patient information and all inference stay inside the institution's network, on hardware it controls.
02 / Interoperable
Standards-based integration with major record systems and bedside devices, on a segmented hospital network.
03 / Clinician-led
Designed with critical-care clinicians, decision support only, with no autonomous clinical action.
04 / Accountable
Full audit trail on every action and inference, with accuracy reported openly, including misses.
See TRIMPACT in your own unit.
Briefings cover the integration surface, the risk models, the audit trail, and the shadow-mode evaluation path.