01 · Cold open
You're the scheduler.
Three units, one week, every shift covered.
What 5:07 changed
- Coverage below target3/4
- Shift starts in1h 53m
- Approaching overtime1
- Unread employee messages3
- Weekend PTO requestPending
- Possible coverage gapSat 3–11
Click to skip
All units at target.
| Employee | Mon | Tue | Wed | Thu | Fri | Sat | Sun |
|---|---|---|---|---|---|---|---|
| Devon King | 11-7 | 11-7 | 11-7 | 11-7 | |||
| Jennifer Lee | 3-11 | 3-11 | 3-11 | ||||
| Lisa M. | 7-3 | 7-3 | 7-3 | 7-3 | 7-3 | ||
| Quentin B. | 7-3 | 11-7 | 3-11 | ||||
| Sara L. | 7-3 | 3-11 | 11-7 | 3-11 | |||
| Tyrell J. | 7-3 | 3-11 | 3-11 | ||||
| Ximena A. | 3-11 | 3-11 | 3-11 | 11-7 | |||
| Yusuf B. | 3-11 | 11-7 | 3-11 | 11-7 | |||
| Zara C. | 11-7 | 3-11 | 7-3 | ||||
| Andre Bell | 7-3 | 7-3 | 7-3 | 7-3 | |||
| Brianna T. | 3-11 | 11-7 | |||||
| Carlos V. | 11-7 | 7-3 | 7-3 | ||||
| Keisha Moore | 3-11 | 3-11 | |||||
| Mike Johnson | 7-3 | 7-3 | 7-3 | 7-3 | 7-3 | ||
| Nadia S. | 3-11 | 3-11 | 7-3 | ||||
| Owen R. | 11-7 | 3-11 | 7-3 | ||||
| Paula K. | 11-7 | 3-11 | 7-3 | ||||
| Tasha Williams | 7-3 | 7-3 | 7-3 | 7-3 | 7-3 |
What 5:07 changed
- Coverage below target3/4
- Shift starts in1h 53m
- Approaching overtime1
- Unread employee messages3
- Weekend PTO requestPending
- Possible coverage gapSat 3–11
Click to skip
One cell went from covered to open. Everything else here is a consequence of that.
01 · The real job
Fourteen moving parts. One person in the middle.
Every day, the scheduler continuously turns an unstable workforce into a safely staffed, compliant building.
Monday was a perfect schedule. Nothing that follows it is unusual.
IllustrativeThe week is a representative script, not a measurement.
Click to skip
A perfect schedule, one week later
- MonStableAll units at target
- TuePTO requestTwo weekend PTO requests land
- WedCall-outLisa M. calls out, Memory Care 7–3
- ThuCensus increaseRehab admits three, requirement +1
- FriOvertime riskTwo CNAs cross 40 hours
- SatTwo uncovered shiftsTwo uncovered shifts: Sat 3–11 Memory Care and Rehab short
- SunNo-show11–7 CNA does not clock in
Monday was a perfect schedule. Nothing that follows it is unusual.
IllustrativeThe week is a representative script, not a measurement.
Click to skip
01 · The little operation
A tiny operating company inside the SNF.
One person is coordinating supply, demand, cost, communication, exceptions, and outside labor.
- Scheduler online
- Facility Schedule×0
- SMS×0
- Phone×0
- Email×0
- Employee availability×0
- Timeclock×0
- Agency portal×0
- Spreadsheet×0
Click to skip
This was one coverage problem.
We digitized the schedule. We never eliminated the coordination work surrounding it.
Eight systems. The scheduler is still the one coordinating between them.
The scheduling system holds the schedule. The scheduler still runs the operation.
The 17-action morning is a representative script, not a measurement.
01 · The call-out, by hand
Handle the call-out.
You're the scheduler. Work it the way it actually happens.
Nothing here resolves instantly, and the clock does not stop while you work.
Your morning, so far
- Messages sent
- 0
- People contacted
- 0
- Systems touched
- 0
- Context switches
- 0
- Time now
- Shift starts in
- 1h 48m
- Coordination minutes
- 0
Facility schedule
One cell is open. Nothing else on this screen knows that.
Internal staff
Hours as of Monday. Availability is whatever was last written down.
Rachel Carter
CoreLast known availability: mornings · 32 hours scheduled
Mike Johnson
CoreMay create overtime40 hours scheduled
Jennifer Lee
PRNAvailability unknown
Cost notes
for the 7–3 shiftNothing here fills itself in. Look it up, write it down.
- Rachel CarterCore teamnot checked · in the timeclock
- Mike JohnsonCore teamnot checked · check Mike's overtime
- Jennifer LeeInternal PRNnot checked · in the timeclock
- Maria RodriguezStaffHealth Familiar Benchnot checked · in the StaffHealth portal
- Vendor B clinicianApproved vendornot checked · in the StaffHealth portal
Group actions
The clock
Another system
Facility policy: try internal first. Unlocks after an internal attempt, or 15 minutes.
Texts
Nothing sent yet. Pick someone.
Outbound messages go out from the staff list and the group actions.
IllustrativeCosts, hours and reply times are simulated from the seeded roster.
Four systems, none of them talking to each other. You are the integration.
02 · The architecture
What if your scheduler didn't have to operate all of this?
Click to skip
Reads from
- Existing scheduling system
- Time & attendance
- HR / payroll
- Employee communication
- External workforce
Decides
Acts through
- Schedule updates
- Employee outreach
- External staffing requests
- Escalation / approvals
- Records / reporting
Operator watches for staffing problems, compares the best options, executes routine recovery workflows, and escalates only when human judgment is needed.
Before: 14 changing variables flow into the scheduler.
After: 14 changing variables flow into Operator. The scheduler gets solutions, recommendations, and true exceptions.
14 changing variables
| Before | After |
|---|---|
| Scheduling system holds the schedule | Scheduling system still holds the schedule |
| Scheduler runs the operation | Operator runs the operation |
| Scheduler handles every transaction | Scheduler handles exceptions and judgment |
02 · The same morning
The same call-out, operated.
The Operator detects it, ranks the options, offers, and confirms.
Click to skip
Memory Care · Wednesday · 7–3
- OpenLisa M.
- CNAQuentin B.
- CNASara L.
- CNATyrell J.
Operator event stream
Recommended recovery plan
Ranking the eligible clinicians.
Six seconds to build the plan. From there, Operator runs the coordination.
Illustrative simulation, not measured customer outcomes.Prototype estimates generated from seeded historical data.
02 · The scenario lab
Change the conditions. The answer changes.
Operator optimizes for the facility.
Conditions
The problem
Your people
Outside
What you value
Try a scenario
IllustrativeIllustrative simulation — not measured customer outcomes.
Who Operator recommends
| # | Candidate | Available | Hours | Cost | Likely to accept | Why? |
|---|---|---|---|---|---|---|
| 1 | Rachel CarterCore team | Yes | 32 → 40 | $176 | 66% | |
| 2 | Maria RodriguezStaffHealth Familiar Bench | Yes | 24 → 32 | $216 | 78% | |
| 3 | Devon KingFloat / sister facility | Yes | 32 → 40 | $176 | 57% | |
| 4 | Sofia AlvarezStaffHealth Familiar Bench | Yes | 16 → 24 | $216 | 73% | |
| 5 | Tasha WilliamsCore team | Yes | 36 → 44 | $2304h OT | 54% | |
| 6 | Mike JohnsonCore team | Yes | 40 → 48 | $2648h OT | 85% | |
| 7 | Grace OkaforStaffHealth Familiar Bench | Yes | 32 → 40 | $216 | 49% | |
| 8 | StaffHealth clinician (network)StaffHealth Network | Yes | 0 → 8 | $224 | 63% |
Offer plan
Opening 1: Rachel Carter 12m → Maria Rodriguez 8m → Devon King 12m
StaffHealth is not favored. It wins only when the conditions make it strongest.
03 · The familiar bench
The best external staffing experience stops feeling external.
External does not have to mean unfamiliar.
- Attendance
- 98%
- Facility familiarity
- High
- Last worked here
- 6 days ago
- Facility favorite
- Yes
- Preferred shift
- 7–3
- Orientation
- Complete
Distance from the center is unfamiliarity, not employment.
Illustrative roster, simulated relationship history.
03 · Operations map
The scheduler is still the workflow engine.
45 recurring workflows. Most still depend on a person to connect the steps.
38human-run workflows
Typical today: 38 human-run workflows. 45 recurring workflows. Most still depend on a person to connect the steps. 7 of 45 are system-assisted. 38 of 45 still require human coordination.
- 7
System-assisted
Runs in today's scheduling software.
- 38
Human coordination
Waits for a person to connect the steps.
45 recurring workflows across the scheduling operation.
Coverage
- Detect open shift · System-assisted
- Detect call-out · Human coordination
- Detect no-show · Human coordination
- Calculate coverage impact · Human coordination
- Restart coverage after no-show · Human coordination
- Update schedule · System-assisted
Availability
- Update availability · System-assisted
- Learn worker preferences · Human coordination
- Predict acceptance · Human coordination
Communication
- Contact employee · Human coordination
- Follow up · Human coordination
- Interpret response · Human coordination
- Broadcast opening · Human coordination
- Notify manager · Human coordination
PTO & Swaps
- Approve routine shift swap · Human coordination
- Analyze PTO impact · Human coordination
- Initiate future coverage search · Human coordination
Cost & Overtime
- Check overtime · System-assisted
- Offer pickup incentive · Human coordination
- Detect recurring overtime · Human coordination
Attendance
- Monitor clock-in · System-assisted
- Detect fatigue risk · Human coordination
- Disciplinary action · Human coordination
Internal Workforce
- Identify internal candidates · Human coordination
- Rank candidates · Human coordination
- Borrow from sister facility · Human coordination
- Detect fairness imbalance · Human coordination
External Workforce
- Escalate to external labor · Human coordination
- Rank StaffHealth familiar bench · Human coordination
- Rank broader StaffHealth network · Human coordination
- Compare other vendors · Human coordination
- Track facility-worker familiarity · Human coordination
Compliance
- Check qualifications · System-assisted
- Check credentials · System-assisted
- Approve high-impact exception · Human coordination
- Override compliance constraint · Human coordination
Prediction
- Detect recurring weekend weakness · Human coordination
- Detect chronic external dependence · Human coordination
- Identify likely FTE deficiency · Human coordination
- Anticipate labor budget pressure · Human coordination
Workforce Planning
- Recommend hiring · Human coordination
- Learn from outcome · Human coordination
- Change staffing policy · Human coordination
Management Reporting
- Produce DON briefing · Human coordination
- Produce administrator briefing · Human coordination
Autonomy is not binary. Even when Operator doesn't make the final decision, it does the work required to reach it.
Illustrative target operating model.
03 · Twelve weeks later
This is no longer a call-out problem.
Same shift. Same unit. Same day. 12 weeks running.
Click to skip
Sat 3–11 · CNA · Memory Care
8 of 12 weeks late
- Week 1: Stable
- Week 2: Unstable
- Week 3: Stable
- Week 4: Unstable
- Week 5: Unstable
- Week 6: Short
- Week 7: Unstable
- Week 8: Short
- Week 9: Short
- Week 10: Short
- Week 11: Short
- Week 12: Short
Week
- Stable
- Unstable
- Short
Saturday 3–11 needed a late fill in 8 of 12 weeks. That is a staffing gap, not a call-out.
Simulated Simulated 12-week history, and the recommendation drawn from it.
03 · Buildable now
This is buildable now.
A few years ago this was much harder. The capabilities matured together.
Why now
Software can finally understand, communicate, decide, and act.
Agentic intelligence
- Understands messy situations
- Reasons across context
- Selects tools
- Plans next actions
- Explains decisions
APIs + tool use
- Reads scheduling, time and HR data
- Calls external systems
- Writes actions back
- Leaves systems of record intact
Voice + text agents
- Conversation becomes software input
- Asks follow-up questions
- Interprets responses
- Triggers structured actions
Live workforce state
- Who, where, when
- Cost and availability
- Reliability and familiarity
- Preferences and history
The breakthrough isn't one technology. It's that these capabilities now work together.
One call-out, end to end
SimulatedClick to skip
Operator checks
- Schedule
- Coverage
- Facility policy
- Overtime
- Available alternatives
- StaffHealth bench
Conversation can now become operational input.
Architecture
Systems of record
Schedule · Timeclock · HRIS · StaffHealth
Live workforce state
Who · Where · When · Cost · Availability · Familiarity · History
Operator
Agent reasoning · Deterministic rules · Optimization · Facility policy
Language model: interprets language, orchestrates tools.
Deterministic rules: credentials, overtime, coverage.
Action layer
SMS · Voice · APIs · StaffHealth · Approvals
Systems updated
Outcome learned · System of record stays authoritative
Autonomy
- Observe
- Recommend
- Execute with approval
- Execute within policy
Technology determines what Operator can do. Policy determines what Operator is allowed to do.
Facility A
Execute with approvalPickup incentives over $50 require approval.
Facility B
Execute within policyPickup incentives up to $100 are pre-approved.
Same intelligence. Different autonomy boundary.
Start with call-out recovery. Expand autonomy workflow by workflow.
The primitives exist. The work is integration, policy, reliability and trust — expanded workflow by workflow.
Architecture sketch, not a build plan.
04 · Every desk
One intelligence layer. Different desired outcomes at every level.
Scheduler. Runs the coordination.
| What this desk watches | Wednesday, 5:07 AM |
|---|---|
| Problems resolved | 1 (Memory Care 7–3) |
| Contacted | Rachel Carter (declined), Maria Rodriguez (accepted) |
| Requires approval | None |
| Remaining exceptions | Saturday 3–11 coverage risk |
Every row is the same Wednesday morning. Only the question changes.
Every desk is reading the same simulated 5:07 AM call-out.
04 · Category context
The market is moving.
Workforce software is heading toward autonomous operations.
Click to skip
Software records work
Schedules, timecards, requests
1 company
OnShift
Senior-care scheduling and workforce software (part of ShiftKey)
What they publish
Publicly positions its platform as software built for senior care, covering recruitment, scheduling, and payroll workflows.
Software recommends actions
Suggested fills, alerts
1 company
IntelyCare
Nurse staffing marketplace
What they publish
Publicly positions its app as a marketplace matching nursing professionals to per diem, contract, and travel shifts at healthcare facilities.
Software executes workflows
Auto-offers, approvals
4 companies
Legion
Workforce management
What they publish
Publicly positions an AI-native workforce-management platform — deterministic, predictive and agentic AI, automated schedule generation, targeted shift offers to fill gaps and cross-location workforce sharing — with non-acute healthcare among its industries.
UKG
Enterprise HCM / workforce management
What they publish
Publicly positions its HCM and workforce-management suite around UKG Bryte AI and agentic workflows, including Dynamic Workforce Operations — automatic call-off and coverage-gap detection with explainable recommendations managers can approve, adjust or automate — plus acuity-based clinical scheduling for healthcare.
ShiftKey
Healthcare shift marketplace
What they publish
Publicly positions itself as a marketplace where licensed healthcare professionals bid on open shifts, with a schedule-automation integration for facilities.
Clipboard Health
Healthcare shift marketplace
What they publish
Publicly positions its marketplace as automatically matching facilities' open shifts with qualified professionals based on credentials and work history.
Operator: scheduler agnostic, a system of intelligence above systems of record
Operator: scheduler agnostic, a system of intelligence above systems of record
Autonomous workforce operations
Policy-bounded agents
1 company
Skypoint
AI data platform for healthcare and senior living
What they publish
Publicly positions itself as an agentic AI platform that unifies data and deploys AI agents to automate operations such as compliance and labor-cost workflows across senior living and healthcare.
Competitors are evidence the category is emerging, not proof the idea is wrong.
Representative public positioning, checked 2026-09-23. Not endorsements.
05 · The demand chain
Today, StaffHealth competes for the demand.
With Operator, it is already inside the system where that demand appears.
| Step | Today | With Operator |
|---|---|---|
| 01 | Facility develops a staffing problem | Staffing problem occurs |
| 02 | Facility attempts internal resolution | Operator sees the demand at creation |
| 03 | A human realizes external staff is needed | Operator attempts optimal internal solutions |
| 04 | A human decides which vendor to use | External labor becomes appropriate |
| 05 | StaffHealth competes for the demand | StaffHealth already exists natively inside the execution environment |
The chains diverge at step two: one waits for a person to decide, one does not.
StaffHealth moves from competing for staffing demand to becoming attached to the creation and resolution of staffing demand.
Escalation order, not a ranking
- 01Core employee
- 02Internal PRN
- 03Float / sister facility
- 04StaffHealth Familiar Bench
- 05Broader StaffHealth Network
Why StaffHealth wins when it wins
- Known pricing
- Integrated workflow
- Known clinician supply
- Facility–worker familiarity data
- Reliability history
- Availability data
- Low friction
- Faster confirmation
- 06Other approved vendors
Illustrative strategy model. The escalation order here is the one the lab's engine actually runs, and nothing in it is weighted for StaffHealth.
05 · The compounding
Operator compounds the business StaffHealth already has.
More revenue from the business we already run. New revenue from the layer we are adding.
Historical demonstrated scale
HistoricalAchieved before national coverage was complete.
StaffHealth operating history, provided. Not computed on this page.
Assumptions
StaffHealth engine
Operator adoption
Permanent hiring / RPO
External vendor marketplace
Enterprise intelligence
StaffHealth engine · reference scale
Not created by Operator1,200 shifts / day × 8 h × 365 × $6 / hour
Where Operator sits
Upstream of the demand, inside the schedule where it is created.
Expand what StaffHealth already does
- Temporary staffing captured through Operator+$4.2M
$21,024,000 reference staffing × 20% captured
- Permanent hiring / RPO+$938K
250 facilities × 25% producing a hire / year × $15,000 per hire
Create new revenue layers
- Operator SaaS+$7.5M
250 facilities × $2,500 / month × 12
- External vendor marketplace+$432K
250 facilities × 40 vendor hrs / month × 12 × $45 × 8% take rate
- Enterprise workforce intelligence+$900K
250 facilities × 30% on the enterprise tier × $1,000 / month × 12
Reference StaffHealth engine
Annual staffing contribution at the selected reference scale.
Operator adds
Incremental value created by Operator
- +$7.5MSaaS·
- +$4.2Mstaffing·
- +$938Khiring·
- +$432Kmarketplace·
- +$900Kenterprise
Illustrative annual economic value to StaffHealth
IllustrativeRevenue and gross contribution shown together for strategic scale, not a single accounting measure.
How to read these numbers
- HistoricalHistorical demonstrated scaleProvided by StaffHealth
- $0.0MIllustrative scenarioAssumptions you can move
- shifts × hours × 365Formula-driven outputRecomputed from the inputs
Customer economics
Today
- Acquire facility
- →Wait for staffing demand
- →Earn staffing contribution
With Operator
- Acquire facility
- →SaaS
- →More staffing demand captured
- →Permanent hiring opportunities
- →Marketplace transactions
- →Enterprise intelligence
One acquisition. Multiple monetization layers.
How it compounds
- Vertical
- More layers on every facility relationship.
- Horizontal
- More of the staffing demand captured at the source.
$13,974,300 Operator adds ÷ 250 facilities
Operator does not require StaffHealth to invent a business from scratch. It compounds a business StaffHealth already proved it can run.
Move any assumption on the left. Every figure beside it recomputes from the formula shown under it.
The reference scale is StaffHealth's own operating history. Every other figure is a formula over the assumptions you set.
05 · The flywheel
StaffHealth distributes Operator.
Operator distributes StaffHealth.
StaffHealth's existing relationships open the door to Operator. Operator opens it back.
The flywheel, as a closed loop of ten steps
- More Operator customers
- More staffing / workforce context
- Better recommendations
- Residual external demand identified earlier
- StaffHealth fills appropriate shifts
- More facility–clinician relationships
- Stronger familiar bench
- Higher fill quality / familiarity
- More valuable Operator
- Greater retention / adoption
- Back to step one
The flywheel
The loop closes on the customer count it started from.
Illustrative model of the loop. No step on this ring is a measured outcome.
06 · 5:07 AM again
This time, the problem never reaches the scheduler.
Good morning.
Overnight: 3 call-outs · 2 coverage risks · 1 overtime issue
- Workforce events
- 6
- Handled automatically
- 5
- Requires judgment
- 1
- Routine coordination performed by scheduler
- 0
$427 potential avoidable labor cost identifiedIllustrative
The one decision waiting on a person
Saturday 3–11 · Memory Care
Recommendation: open to Familiar Bench now
Why: internal availability is thin and delay increases fill risk
| Employee | Mon | Tue | Wed | Thu | Fri | Sat | Sun |
|---|---|---|---|---|---|---|---|
| Maria R.StaffHealth Familiar Bench | 7-3 | ||||||
| Devon King | 11-7 | 11-7 | 11-7 | 11-7 | |||
| Jennifer Lee | 3-11 | 3-11 | 3-11 | ||||
| Quentin B. | 7-3 | 11-7 | 3-11 | ||||
| Sara L. | 7-3 | 3-11 | 11-7 | 3-11 | |||
| Tyrell J. | 7-3 | 3-11 | 3-11 | ||||
| Ximena A. | 3-11 | 3-11 | 3-11 | 11-7 | |||
| Yusuf B. | 3-11 | 11-7 | 3-11 | 11-7 | |||
| Zara C. | 11-7 | 3-11 | 7-3 | ||||
| Andre Bell | 7-3 | 7-3 | 7-3 | 7-3 | |||
| Brianna T. | 3-11 | 11-7 | |||||
| Carlos V. | 11-7 | 7-3 | 7-3 | ||||
| Keisha Moore | 3-11 | 3-11 | |||||
| Lisa M. | 7-3 | 7-3 | 7-3 | 7-3 | |||
| Mike Johnson | 7-3 | 7-3 | 7-3 | 7-3 | 7-3 | ||
| Nadia S. | 3-11 | 3-11 | 7-3 | ||||
| Owen R. | 11-7 | 3-11 | 7-3 | ||||
| Paula K. | 11-7 | 3-11 | 7-3 | ||||
| Tasha Williams | 7-3 | 7-3 | 7-3 | 7-3 | 7-3 |
Filled by Operator · 5:21 AM
Illustrative simulation, not measured customer outcomes.
Click to skip