Operational efficiency for healthcare
Better care begins
in the back office.
We believe the way to improve care is to help the people who deliver it get more from every hour and every dollar. Our mission is to make healthcare agencies operationally efficient.
01Problem
Care is capped by administration.
Demand for care is growing and the operations behind it have not kept up.
Every new client adds scheduling, compliance, and billing work, so admin scales proportionally with client growth. Agencies hit a ceiling set by coordinator capacity.
The industry calls this overhead. We think it is the central problem. Solving it has a huge impact on the agency's business, and we believe that is the condition for better health outcomes for patients.
02Why We're Building This
Our Thesis.
Better care begins in the schedule, the claim, and the payroll run. That is the part of home care still waiting to be built, and it is where we work.
-
Care is delivered through operations.
Every visit is set up before it happens. Someone builds the schedule, assigns the caregiver, and passes along what she needs to know. Those decisions determine who arrives, how long they stay, and how much of the visit goes to the client. Improve the decisions and the care improves with them.
-
Efficiency turns into wages.
Home care runs on single digit margins. Money lost to rework, denied claims, and unplanned overtime is money that could have gone to pay, training, and unhurried visits. Recovering it is how an agency affords to do the work well.
-
Turnover is an operating number.
Caregivers leave over short paychecks, Tuesdays that change on Monday night, and an office that goes to voicemail. All three are built in the back office, which means retention responds to the same fixes as everything else there. Most of the industry treats it as weather instead.
-
This layer was skipped.
Software has already rebuilt every industry that schedules people and moves money, most of them twice. Home care does both, under clinical and regulatory load, and still runs on faxes, binders, and a spreadsheet with a version number in its name. Caregivers and families carry the cost of that delay.
-
Scale should be a choice.
A forty caregiver agency answers to the same payers, the same regulators, and the same families as a national operator. The national operator has a department for it. We put that department in the software, so an owner grows when she decides to.
03Work
One system for the entire back office.
We build a single operational system for care agencies — deliberately narrow and deep. Everything a back office must do, in one place, designed as one motion rather than a dozen vendors stitched together. We are building it first for home care: the fastest-growing and least equipped setting in healthcare.
- Scheduling & matching
- Coverage decided in minutes, not phone trees.
- Visit verification
- EVV, Cures-Act compliant, aggregator-ready.
- Care documentation
- Notes, plans, and attestations in one reviewable record.
- Claims & billing
- Medicaid authorization through 837P submission.
- Payroll
- Verified visits become payable hours, automatically.
- Compliance
- Credentials, incidents, and audit trails kept current.
04Applied AI
Agents can’t run a business they can’t read.
We are Silicon Valley engineers by training and applied-AI people by craft — careers spent making AI do real work inside real institutions, not demos of it. We brought that craft to healthcare because no industry is owed it more.
It taught us an unfashionable discipline: AI fails when it is sprinkled on top of a business it cannot read. So we work with each agency in a fixed order — the data, then the ontology, then the agents. Never the reverse.
-
Step I · Unify
One uniform dataset.
Together we define the agency’s uniform dataset and its sources — schedules, visit records, authorizations, claims, credentials, payroll — and resolve them to a single source of truth.
-
Step II · Model
An ontology of the business.
On that foundation we build the agency’s ontology: clients, caregivers, visits, claims, and the rules that bind them — expressed in terms a machine can reason over and an auditor can inspect.
-
Step III · Automate
Agents do the work.
Only then do AI agents take over the repetitive weight of the back office — reconciling, drafting, flagging, filing — while people review, approve, and remain unmistakably in charge.
- 07:58agentreconciled 41 verified visits into payroll hours
- 08:03agentdrafted an 837P claim, six lines, queued for review
- 08:11agentflagged a credential expiring in 14 days, renewal sent
- 08:19agentmatched Saturday’s open shift, three caregivers notified
- 08:24ownerreviewed and approved — the only human minutes spent
05Company
Independent, and deliberately close to the work.
Hearthloop is an independent, founder-led company. We keep the team small and the standards high, and we work directly with the agencies we serve — the person who answers your email is the person who builds the product.
We measure ourselves in our customers’ terms: hours returned to the week, dollars returned to care, and agencies that grew because their operations finally could.