Patient categorization
CORA predicts each patient's physical care demand from three standard measures already collected for skilled home health patients, using a peer-reviewed model co-authored by our founder at Ohio State.
Injury-aware scheduling support
Category-Optimized Resource Assignment
You cannot extend a workforce you are quietly wearing out. CORA protects the bodies doing the work.
Home health aides perform some of the most physically demanding work in health care, alone, in rooms never designed for care. CORA is AI decision support that matches assignments to the physical demand of each patient, so the heaviest caseloads stop landing on whoever happens to be available.
The reality
Bathing, transfers, and mobility assistance are performed alone, in confined residential bathrooms, with equipment that was designed for institutions and does not fit the home. Those tasks drive a large share of the musculoskeletal injuries aides suffer.
Scheduling has to balance geography, availability, continuity, and patient needs, and almost none of it accounts for how physically heavy a caseload has become. So high-demand assignments quietly pile onto the same people, week after week, until a body gives out.
The dominant scheduling platforms match on availability, credentials, proximity, and continuity. None of them match on physical demand. That is the gap CORA fills.
What CORA does
CORA predicts each patient's physical care demand from three standard measures already collected for skilled home health patients, using a peer-reviewed model co-authored by our founder at Ohio State.
Deterministic rules apply the published exposure thresholds rather than leaving them to memory. Thresholds and constraints are agency-configurable, with the published values as defaults, and every deviation is logged.
When an open visit needs a name, CORA explains the physical-demand consequence of each candidate assignment in natural language a scheduler can act on immediately.
CORA tracks the load building across each aide's actual week, not a generic capacity, and flags when someone is drifting into elevated risk.
It can run standalone or layer into the EMR or scheduling platform an agency already uses, starting with a low-risk CSV import and moving to live connectors aligned to FHIR.
Deterministic safety rules, human-in-the-loop overrides, and logged bias auditing are part of the architecture rather than a policy bolted on afterward.
How it works
Each patient is scored for physical demand from that patient's own capabilities, using measures the agency already collects.
CORA sits inside the moment the schedule is being built and shows what each possible assignment would cost the caregiver's week.
The human scheduler remains responsible for the final decision. CORA reduces the friction, never the responsibility.
Who it serves
Home health aides
Caseloads that account for the physical toll of the work, and careers that can last without an injury ending them.
Schedulers
The physical-demand consequence of each assignment, surfaced at the exact moment an open visit needs a name.
Agency leadership
Fewer injuries, lower workers' compensation exposure, and a defensible, auditable record of how assignment decisions were made.
Patients aging in place
Continuity of care from caregivers who are still able to show up next month.
In their words
“It's a difference maker for the agencies that care, and everybody should.”
“I like it a lot for the 'I'm not going to overload you out of the gate and you're going to quit in three months, because you're tired and you're stiff and you're sore.'”
“If you have three options there on your open visits, send something to all three… no PHI… It alleviates Teresa having to make three phone calls and then show me more people and make six phone calls. It's just offer, offer, offer, offer, and send it out.”
“I think it's great. I do. I need everybody to use this.”
What we are measuring
Where this comes from
CORA is grounded in research co-authored by Laura Czuba at Ohio State, which showed patient physical care demands can be predicted from three standard measures, and found an odds ratio of 6.4 for aide-reported pain on days where half of visits involved the highest-demand patients.
The categorization-informed scheduling approach was piloted manually at a multi-site home care organization before any software existed. CORA is that practice, made consistent and explainable.
CORA is at TRL 3 with a working proof-of-concept. Figures cited draw on BLS 2024, PHI 2025, Czuba et al. 2012, King et al. 2018 and 2019, NIOSH 2012, and related published research. This page describes a product in development, not a shipped commercial system.
Tell us who you are and we will reach out when there is something real to try.