Platform
The operating platform being built for Hospital-at-Home teams.
We are designing the platform for lean hospital teams — not only large academic programs with dedicated departments. Two nurses, a director who also runs case management, and a CFO who needs the numbers by Thursday.
Surface more potential candidates
The first release is being designed to analyze relevant ED, observation, and inpatient data and surface potential candidates before the opportunity is missed. Each candidate will be presented with the supporting information used by the model, and the hospital's clinicians will make the final decision.
- Planned screening across connected ED, observation, and inpatient data
- Supporting clinical and operational context visible for review
- Structured acceptance and decline reasons to improve future evaluation
Measures: candidates surfaced per day — the top of the funnel every stalled program is missing.
Community-acquired pneumonia
Meets criteria: CURB-65 of 1, room air sat 95%, lives with spouse at home, within service radius.
Cellulitis, lower extremity
Meets criteria: afebrile 18h, IV antibiotics only, no ICU history. Awaiting social work note.
CHF exacerbation
Borderline: diuresing well, but lives alone — needs caregiver confirmation.
COPD exacerbation
Meets criteria: off BiPAP 24h, ambulatory, home O₂ already in place.
Bring the day into one prioritized workspace
The platform is being designed to combine device signals, patient messages, visit status, and operational tasks in one workspace. Signals will be organized using hospital-approved escalation logic, with ownership and supporting context visible to the care team.
- Severity tiers, each with a documented escalation path and an owner
- Non-clinical noise identified for review rather than silently suppressed
- Visit routing designed to account for drive time and traffic, not straight-line distance
Measures: alert volume, actionable-alert rate, response time, and unresolved work.
SpO₂ 88% sustained 6 min, no response to call
2 min ago · RN Alvarez notified
Missed evening visit — no answer at door
14 min ago · assigned to on-call
HR trending up 12 bpm over 4h, afebrile
38 min ago · queued for rounds
Patient message: question about antibiotic timing
1h ago
Cuff disconnected — reseated by caregiver, confirmed
2h ago · closed by RN Okafor
Prove the value
Your CFO has about eight questions, and none of them are about software. The reporting layer is being designed to connect operational performance with the financial and quality measures hospital leaders need to evaluate the program. Intended to be exportable and useful in a finance committee.
- Contribution margin per episode, designed to be measured against the inpatient alternative
- Length of stay, escalation rate, readmissions, and patient experience designed for one view
- Payer mix and waiver compliance reporting designed as a single export
Measures: contribution margin per episode — a number that keeps the program funded.
Average daily census
11.4
+2.1 vs last month
Contribution margin / episode
$3,180
+$240
Average length of stay
4.1 d
−0.3 d
Escalation to inpatient
6.2%
−1.4 pts
30-day readmission
9.1%
−2.0 pts
Patient experience
4.8 / 5
n = 212
Referral conversion
63%
+9 pts
Days to break-even
Month 7
1 ahead of plan
Under the hood
What the first release is being designed to do.
Four properties that will matter when clinical AI is evaluated in real workflows.
Planned continuous ingestion
HL7v2 and FHIR feeds from your EHR, designed to be normalized into one patient timeline. The intended direction is streaming rather than a nightly batch — subject to integration, testing, and hospital approval.
Evidence-first inference
Model outputs are being designed to include the supporting factors used to produce them. A clinician should see "CURB-65 of 1, room air sat 95%, spouse at home" rather than a score with no provenance. The initial product is not intended to independently diagnose or treat patients.
Feedback as training data
Accepts, declines, dismissals, escalations, and outcomes are intended to be captured as evaluation signals by the same system designed to support the program. Model performance can then be assessed against the feedback from the people using it.
Planned network-level improvement
Future model updates may be evaluated across the founding-partner network. Patient records would remain within the applicable hospital environment, subject to the agreements and governance that are established.
Trust
Designed for hospital-grade trust from the beginning.
Tenant isolation
The platform is being designed around tenant isolation and hospital control of patient data. Before production deployment, the precise data architecture, permitted model uses, retention terms, and opt-out rights will be documented contractually with each hospital.
BAAs before anything connects
A Business Associate Agreement will be executed before the platform receives or connects to protected health information. Access is designed to be role-scoped, least-privilege by default, and reviewable by your compliance team.
Clinicians always decide
The initial product is not intended to independently diagnose or treat patients. It is designed to surface information and organize work for review by licensed hospital clinicians. Recommendations are intended to show the supporting data rather than present a black box.
Audit trails you can actually read
Production deployments are being designed to include role-based access controls and readable audit logs for access, alert actions, and workflow decisions.
See the platform taking shape.
Or start with the Opportunity Index preview — it takes 30 seconds. Try the preview
