Solutions · Hospital Systems
See the whole health system operating at once.
Claritus.One connects operational signals across facilities and systems, identifies what requires attention, explains why it matters, and coordinates the response across teams and AI agents.
Health system · all facilities
Enterprise
- Hospitals
- 8
- Outpatient
- 42
- Regions
- 3
- Med-surg · 1 PM
- 97%
- target 92%
- ED boarders · 3 PM
- 9
- projected
- Discharges by noon
- 6 / 14
- on current pace
Signals · Central Medical Center
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
The operating problem
The problem is not the absence of systems. It is that decisions happen between them.
A health system runs on dozens of applications, and each one accurately represents its part of the organization. Very few of them understand the operation as a whole — so that understanding lives in phone calls, huddles and spreadsheets.
- EHR
- Encounters, orders, ADT events, clinical milestones
- ERP
- Supply, finance, cost centers, purchasing
- Workforce management
- Schedules, time, skills, overtime, open shifts
- Scheduling
- Templates, appointment inventory, provider calendars
- Patient access
- Registration, referrals, authorizations
- Transfer center
- Inbound requests, acceptances, declines
- Bed management
- Bed status, cleaning queue, assignments
- Revenue cycle
- Claims, denials, work queues, payments
- Call center
- Call volume, wait time, abandonment
- CRM
- Outreach, campaigns, patient engagement
- Analytics platforms
- Dashboards and the metric definitions behind them
- Data warehouse
- Modeled history across every source
- Spreadsheets
- The trackers your team actually runs the day on
- Departmental applications
- One department, in great detail
The core idea
Your hospital already has systems of record. Claritus.One becomes its system of operational intelligence.
The intelligence and orchestration layer across hospital operations: it reads what every system already knows, understands how those facts relate across departments and facilities, and coordinates what happens next.
Connected operations
Hospital operations don’t happen in silos. Your intelligence shouldn’t either.
A hospital is not a collection of independent departments. A change in one propagates into the others, usually faster than the reporting that would reveal it. Choose where the change enters and follow it through the building.
Reaches every domain shown
- Emergency Dept+0
ED arrivals
142
arrivals by noon
Waiting on upstream change…
- Admissions+40 min
Inpatient demand
31
admits from ED
Waiting on upstream change…
- Capacity+1 h
Med-surg bed demand
91%
occupancy at 1 PM
Waiting on upstream change…
- Nursing+1 h
Nursing coverage
1.00
coverage, 5B & 6A
Waiting on upstream change…
- Environmental Svcs+2 h
Bed turnaround load
18
turnarounds by 3 PM
Waiting on upstream change…
- Case Management+2 h
Discharge timing
9
needed by 1 PM
Waiting on upstream change…
- Transfer Center+3 h
Transfer capacity
6
beds open to transfers
Waiting on upstream change…
- Perioperative+5 h
Elective capacity
0
elective admits at risk
Waiting on upstream change…
- Emergency Dept+4 h
Diversion risk
Low
ED diversion risk
Waiting on upstream change…
What Claritus.One sees
Each signal is inside its own tolerance. Together they describe an afternoon capacity constraint: med-surg at 97% by 1 PM, transfer acceptance down to two beds, and diversion risk elevated by 4 PM.
Response routed to
- Patient Flow Agent — prioritize 14 high-probability morning discharges
- Staffing Agent — float options for 5B and 6A, awaiting the house supervisor
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
The enterprise operating picture
One intelligence layer across every operating domain.
Six domains, one model of the organization underneath them. The value is not any single domain — it is that a staffing question and a capacity question are finally answered from the same facts.
Claritus.One · the intelligence layer
One model of the organization · every domain reads and writes through it
Connections · Patient Flow
What changes in patient flow rarely stays there. These are the dependencies Claritus.One reasons across when it decides what deserves attention.
- Patient FlowWorkforceEvery bed opened is a nursing assignment. Capacity that cannot be staffed is not capacity.
- Patient FlowRevenue CycleAvoidable days and late discharges show up later as length-of-stay variance and denials.
- Patient FlowEnterprise ServicesThe transfer center can only accept what inpatient capacity can absorb this afternoon.
- Patient FlowOperational PerformanceThroughput is the denominator behind most facility and service-line variance.
Beyond the command center
A command center shows you the operation. Claritus.One helps operate it.
Most health systems already have dashboards, analytics teams, capacity tools and a daily operating call. Claritus.One does not replace them with another screen. It moves the organization up the ladder from what happened to what happens next.
Stage 1 · What happened?
Traditional analytics
Metrics and dashboards. Every number is correct, each one is on its own tile, and connecting them is left to whoever is on the morning call.
- ED arrivals
- +11%
- vs expected, 6–7 AM
- Discharges before noon
- 9
- plan: 23
- Med-surg occupancy
- 92%
- current
- 5B RN hours
- −6.0
- vs grid, 3–11 PM
- 6A RN hours
- −4.0
- vs grid, 3–11 PM
- Transfer requests
- 4
- pending
Six correct numbers, owned by four departments. Nothing on this screen says they are the same problem — that conclusion is reached on the 9 AM call, if it is reached at all.
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
Enterprise Daily Brief
Before the 9 AM operating call, the system has already done the triage.
One brief for the whole enterprise, written in sentences. Open any item to its evidence, ask why it was raised, inspect the units behind it — or turn the recommendation into a workflow without leaving the page.
Good morning.
Across the system, inpatient capacity is stable, but three operational issues deserve attention before 9 AM.
The first is time-sensitive: Central Medical Center is heading into a tight afternoon, and the discharges that would prevent it need to start this morning. The other two are patterns rather than events, and both have a decision attached.
System picture
- Hospitals operating normally
- 6 of 8
- Need a decision before 9 AM
- 2
- Patterns being watched
- 1
- Agents working now
- 5
- Closed overnight
- 3
What the narrative is built on
- Med-surg occupancy, system
- 0%
- Target 92%
- ED arrivals vs expected
- +0.0%
- CMC +11%
- Discharges before noon
- 0
- Plan 74
- Open shifts, next 72 h
- 0
- −9 since yesterday
- Call abandonment
- 0.0%
- Cardiology 14.2%
- Denial work queue
- 0
- Within capacity
Needs attention · in priority order
3 of 212 observations raised · 209 suppressed as routine
- Likely impact
- Medical-surgical capacity could tighten beginning around 1 PM.
- Claritus recommendation
- Prioritize 14 high-probability morning discharges and review two units with delayed discharge milestones.
- EHR · ADT62 ED arrivals since midnight against an expected 56 for a Tuesday in this season.
- EHR · OrdersDischarge orders written before 9 AM are at 9, against a 30-day median of 17.
- Capacity modelProjected med-surg occupancy reaches 97% at 1 PM against a 92% operating target.
- EHR · MilestonesUnits 5B and 6A hold 9 of the 14 patients with open discharge dependencies.
Suggested owner: House Supervisor, CMC
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
The hospital operations graph
Claritus.One doesn’t just ingest healthcare data. It understands operational context.
Underneath every observation is a model of how the hospital actually works: which demand lands on which beds, which beds wait on which discharges, which units need which nurses. Select anything to see what it depends on and what depends on it.
Upstream: what it depends on · Downstream: what it affects
Demand
Capacity
Workforce
Access & Revenue
Unit · 5B Med-Surg
Unit capacity
32 of 32 by 2 PM
Where the beds actually are. Capacity is a unit-level fact before it is a hospital-level one.
Depends on
Discharge readiness frees Unit capacity
Each completed discharge frees a bed on a specific unit — 5B has five ready patients blocked on open milestones.
Affects
Unit capacity requires Nursing requirements
A full 5B at today’s acuity requires eight RNs on day shift. Six are scheduled.
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
Multi-facility intelligence
Start with the enterprise. Get to the reason in seconds.
Enterprise visibility usually costs local context, and local context usually costs the enterprise view. Move from the system to a single unit’s open discharge dependencies without losing where you came from.
Level 1 of 5 · System
Enterprise
8 hospitals, 42 outpatient locations, 3 regions. 2 hospitals require attention.
Bars show projected med-surg occupancy at 1 PM · rule marks the 92% operating target
AI Workforce
Not assistants. Controlled operational workers.
Each agent has a defined role, the systems it may read, the actions it may take, and the point at which it must stop and ask. They work continuously in the background — an additional operating layer, not a chat window.
Patient Flow Agent
Agent workingContinuously evaluates flow conditions, identifies emerging constraints, investigates contributing factors and coordinates defined workflows.
Objective · Keep projected demand inside available capacity, hours before the constraint arrives.
- Reads
- EHR · ADT events and discharge milestones
- Bed management · status and cleaning queue
- Transfer center · inbound requests
- May do
- Rank discharge candidates by probability and dependency
- Route discharge priority lists to unit leaders
- Open and track flow exceptions
- Requires approval
- Changes to elective admission timing
- Requests to divert or decline transfers
Escalates when · Projected med-surg occupancy above 95% for more than two hours escalates to the house supervisor.
Working now
212 tasks today · Routing only · no approvals required
Afternoon med-surg capacity at Central Medical Center
- Found
ED arrivals tracking 11% above the expected morning pattern
Projected inpatient demand of 23 admissions by 3 PM against 17 expected.
- Found
Projected discharges below the level required to absorb it
9 discharges completed by noon on a typical Tuesday; 4 are on track today.
- Found
14 patients have high discharge probability with open dependencies
Pending transport, pharmacy reconciliation and two outstanding consults.
- Did
Routed a prioritized discharge list to 5B and 6A unit leaders
- Waiting
Tracking milestone progress against the 1 PM threshold
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
Controlled autonomy
AI should have authority. Not unlimited authority.
Governed autonomy, not “give AI access to everything and hope for the best.” Authority is granted one level at a time, per agent and per class of work, and can be withdrawn at any level.
Authority is granted per class of work, not per agent
Swipe to see all six levels →
Float within same-skill pool, no overtime
Ceiling · Execute
Assignments that add overtime
Ceiling · Prepare
Cross-facility moves
Ceiling · Recommend
Agency or incentive pay
Ceiling · Investigate
Level 05 · Execute
- May
- Carry out work in classes that policy has pre-approved, inside thresholds.
- May not
- Exceed a financial threshold, scope limit or policy — that drops it back to Prepare.
Staffing Agent · Publishes the 4C → 5B float: same skill, no overtime, pre-approved class.
Every agent operates within
- Defined roles
- Permissions
- Policies
- Data-access boundaries
- Financial thresholds
- Approval requirements
- Escalation rules
- Human oversight
- Complete activity history
Illustrative policy configuration. Thresholds and ceilings are set by each health system.
Human + AI operations
More operating capacity at every level of the organization.
Claritus.One does not replace your operations teams. It gives them a layer that does the watching, the first investigation and the coordination — so the people with judgment spend it where it counts.
AI Workforce
Patient Flow · Staffing agents
Detects a capacity risk and performs the initial investigation
Connects rising ED arrivals, lagging discharges and two units near staffing thresholds at Central Medical Center. Assembles the evidence and two response options.
Agent · 4 min
Coordinates the approved actions
Routes discharge lists to unit leaders, chases transport and pharmacy dependencies, publishes the float, and re-runs the 6A check at 10 AM as instructed.
Agent · continuous
Department leadership
Med-surg nursing director, CMC
Reviews the recommendation
Approves prioritizing 14 discharges and one float move from 4C to 5B. Declines the second float and asks for 6A to be re-checked at 10 AM.
Human · 6 min
Enterprise operations
System capacity command
Tracks the outcome
11 of 14 discharges complete. Projected med-surg occupancy back to 91%. No diversion, no overtime added. Closed on the system view with its full history.
Human · 2 min
AI Workforce
01 · 7:26 AMDetects a capacity risk and performs the initial investigation
Connects rising ED arrivals, lagging discharges and two units near staffing thresholds at Central Medical Center. Assembles the evidence and two response options.
Agent · 4 min
Department leadership
02 · 7:41 AMReviews the recommendation
Approves prioritizing 14 discharges and one float move from 4C to 5B. Declines the second float and asks for 6A to be re-checked at 10 AM.
Human · 6 min
AI Workforce
03 · 7:42 – 11:30 AMCoordinates the approved actions
Routes discharge lists to unit leaders, chases transport and pharmacy dependencies, publishes the float, and re-runs the 6A check at 10 AM as instructed.
Agent · continuous
Enterprise operations
04 · 12:40 PMTracks the outcome
11 of 14 discharges complete. Projected med-surg occupancy back to 91%. No diversion, no overtime added. Closed on the system view with its full history.
Human · 2 min
Human time spent: 8 minutes. Agent work: continuous, from 7:26 AM to close.
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
Operating scenarios
Four mornings, four domains, one way of working.
Each scenario runs the same sequence — the problem, what Claritus sees, what it finds, what it recommends and who does the work. The domain changes. The operating model does not.
Expected admissions vs. discharges, per hour
Open med-surg beds reach zero ≈ 1 PM
- The problem
Demand is rising while discharge activity is falling behind.
ED arrivals are running 11% above the morning pattern. Discharges are tracking the usual curve — which is now too slow for the day the hospital is actually having.
- Claritus sees
Several independent signals together indicate an afternoon capacity constraint.
ED arrivals, admission conversion, discharge probability and two units near staffing thresholds. None crosses its own alert line. Together, med-surg runs out of beds around 1 PM.
- Claritus investigates
Discharge progression is deviating on two units, not across the hospital.
On 5B and 6A, discharge orders are written on time but transport, pharmacy and family pickup milestones are running 2–3 hours late. Fourteen patients have a high probability of leaving before noon.
- Claritus recommends
Prioritize 14 morning discharges and clear their open dependencies.
A ranked list by unit with the specific blocker on each patient, plus a review of two elective admissions scheduled into the 2 PM window.
- Claritus coordinates
Routed to unit leaders and the house supervisor.
The Patient Flow Agent posts the list to the 10 AM unit huddles, requests transport for six patients and tracks every milestone through to the bed being released.
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
Interoperability
Your systems don’t need replacing. They need an intelligence layer between them.
No migration and no new system of record. Claritus.One sits above what you run, builds one operational understanding from all of it, and sends decisions and actions back into the workflows your teams already use.
Read continuously from every system — nothing migrated, nothing replaced
Claritus.One
- Intelligence
- Operations
- AI Workforce
Operational context
One resolved model of the health system
- Facilities
- Units
- Beds
- Staff
- Schedules
- Patients in flow
- Queues
EHR
ADT events, orders, discharge milestones
ERP
Cost centers, supply, purchasing
Workforce management
Schedules, time, skills, open shifts
Scheduling
Templates, appointment inventory
Patient access
Registrations, referrals, authorizations
Revenue cycle
Claims, denials, work queues
CRM
Outreach, engagement, campaigns
Telephony
Call volume, wait time, abandonment
Data warehouse
Modeled history across sources
Analytics
Dashboards and metric definitions
Departmental systems
One department, in detail
Existing health system · stays the system of record
Patient Flow → 10 AM unit huddle
Ranked discharge list with the blocker on each patient
Workforce → Staffing office
Two float assignments, prepared and awaiting approval
Patient Access → Contact center
Callback queue prioritized by referral age
System categories shown generically. No vendor relationships are implied.
Integration catalog
Every system category, what Claritus reads, and what it sends back.
Whatever your system runs — Epic or Oracle Health, UKG or symplr — Claritus reads it through standards-based feeds, vendor APIs and warehouse reads. Systems of record stay authoritative; actions go back only where a workflow lives.
ReadsWrites back
Claritus reads
- Encounters and orders
- Discharge orders and milestones
- ED arrival and disposition timestamps
- Consult and imaging status
Writes back
- Discharge-dependency tasks to care teams
- Unit huddle worklists
Commonly run here
EpicOracle HealthMEDITECHAltera Digital HealthTruBridge
Connects via
- FHIR APIs
- HL7 v2 feeds
- Warehouse reads
Feeds
- Patient Flow
- Operational Performance
Clinical & capacity
EHR
Claritus reads
- Encounters and orders
- Discharge orders and milestones
- ED arrival and disposition timestamps
- Consult and imaging status
Writes back
- Discharge-dependency tasks to care teams
- Unit huddle worklists
Commonly run here
EpicOracle HealthMEDITECHAltera Digital HealthTruBridge
Connects via
- FHIR APIs
- HL7 v2 feeds
- Warehouse reads
Feeds
- Patient Flow
- Operational Performance
15 system categories · 9 receive actions back · Epic, Oracle Health, MEDITECH, Altera Digital Health, TruBridge, InterSystems, Rhapsody, NextGen Connect (Mirth), TeleTracking, Care Logistics, Central Logic, UKG, symplr, QGenda, ADP, Workday, Oracle, Experian Health, Genesys, NICE, Five9, Cisco, Salesforce, Microsoft Dynamics 365, Waystar, Infor, SAP, Snowflake, Databricks, Microsoft Fabric, Tableau, Power BI, Microsoft Excel, Google Sheets, Smartsheet are trademarks of their respective owners.
How integration worksEnterprise architecture
A healthcare intelligence cloud, above the systems you already run.
Intelligence and the AI Workforce sit on a shared operations layer, which sits on one operational context, which is built from your existing systems. Every layer runs the same closed loop.
Operational Context
One resolved model of how the health system actually works.
Facilities, units, beds, staff, shifts, schedules, patients in flow and work queues — and the relationships between them. This is what lets a staffing answer and a capacity answer agree.
Select any layer to see what it does
The closed loop
Observe. Understand. Decide. Act. Measure. Learn.
Dashboards stop at the first stage. Claritus.One runs all six, continuously, and closes the loop by measuring whether what it did actually worked.
Observe · 7:14 AM
ED arrivals at Central Medical Center are running 11% above the expected morning pattern.
Read continuously from the EHR, alongside discharge milestones, staffing and transfer requests across all eight hospitals.
1.84M
signals read overnight
Illustrative demonstration data for a fictional eight-hospital system. Not a customer, and not customer results.
Enterprise governance
Designed for governed enterprise operations.
The controls an enterprise expects of anything that can act inside its operation. Your security, privacy and compliance teams will have a longer list, and that is a conversation we have directly.
Role-based access
What a person can see follows what they are responsible for. A unit manager sees their unit; a regional president sees the region.
Organization hierarchy
System, region, hospital, department and unit are modeled as the organization actually reports — not flattened into a list of locations.
Facility-level permissions
Access, agent authority and approval rules can differ by facility, because a critical access hospital and a flagship are not run the same way.
Agent permissions
Read and write are separate grants, per system and per action. Nothing an agent can do is implied by something else it can do.
Approval policies
Thresholds — financial, clinical-adjacent, scope of change — decide when work stops and waits for a named person.
Activity history
Every observation, recommendation, approval, override and action is recorded with the evidence that was in front of it at the time.
Auditability
Any outcome can be traced back to the signal that started it, the reasoning applied, and the human or agent that acted.
Human escalation
Escalation paths are defined per workflow, with timeouts, so nothing an agent cannot resolve waits silently.
Workflow governance
Workflows are versioned, owned, and changed deliberately. What an agent did last month can be explained against the rules in force last month.
One platform
Two operating stories. One underlying architecture.
Claritus.One is not one product for urgent care and another for hospitals. Both are the same intelligence, operations, AI workforce and data layer — shaped to a different organization.
Metaphor · Network
Urgent Care
Run a distributed network of locations as one intelligent operation.
Dozens to hundreds of centers · fast, high-variability decisions
See Urgent CareMetaphor · Operational graph
You are hereHospital Systems
Connect a complex healthcare enterprise into one intelligent operating environment.
Facilities, departments and units · deeply interconnected operations
One Claritus.One platform
The same architecture underneath both. What differs is the shape of the organization it is modeling.
Request a demo
Your health system already generates the signals. Claritus.One turns them into coordinated action.
Bring the operating question that costs your system the most — a capacity afternoon, a staffing pattern, a growing work queue. We will show you how Claritus.One would have surfaced it, explained it and coordinated the response, across an enterprise built to look like yours.