Straight answers, before you talk to anyone.

Book your call. One 30-minute call, then a working preview in your brand within 24 hours, loaded with sample people and sample notes. A self-serve path for smaller organizations is coming; today every preview starts with a call.

Nothing. The preview is complimentary; we call the tier Preview Edition. No real health information is entered during the preview, and nobody signs anything before your team has tried it.

The people who will actually use it. Agencies: a scheduler, a nurse, and a family member if you can. Community and day programs: one week of your current closeout and your intake form. Health systems: whoever owns the program and whoever will ask the security questions.

About ninety days, with our team beside you. Days 1 to 30: your brand, setup, guardrails, and staff sign-on scoped with your IT team. Days 31 to 60: your existing PDFs, videos, and FAQs become check-ins and lessons with your reviewers' approval. Days 61 to 90: training, a soft launch with one program, then rollout.

Your logo, your colors, and your words. Then the PDFs, videos, and FAQs you already hand to families. We turn them into check-ins and lessons, and your reviewers approve each one before it goes out.

Families need nothing to install: text, email, a code to scan, or a link to tap. Your staff use a web console with multi-factor sign-in. Staff single sign-on and any connection to your records system are scoped with your IT team during onboarding; no partner runs single sign-on today.

In the third month, alongside a soft launch with one program. Your team's console holds a training library: branching simulations, short huddles, activity plans, and courses that run again with every new hire. Each learner can export a certificate. Planned, not shipped: an organization-wide training roll-up for managers.

Neither. Any community program gets a warm front door by code, link, or shared tablet, and public resource pages under your name that need no login. Reminders, outreach, and volunteer coordination come with it. Day programs add one workflow from intake to an evening note. GUIDE partners are welcome; GUIDE is not required.

Your name and colors, not ours. If you turn on the family loop, they send a few words at drop-off in the morning and read an evening note that night. The note is drafted from what actually happened that day, in the style they chose at intake: warm, brief, or clinical. Reservation requests wait for your staff to approve, waitlist, or decline.

A person on your staff approves every evening note before it goes home. The AI drafts it from what actually happened that day, in four short sections: To talk about tonight, Ideas for home, Caring for yourself, For your next visit. It cannot add a fact that is not in the day's record.

Less, or we recommend pausing. Select everyone at the table, one choice, one button that reads Save for 14. The emotion dial includes Unable to tell, which stores nothing. A 3:45 readiness check flags gaps while staff can still fix them. We recommend pausing a rollout if the software adds duplicate after-hours work, and we build the reports so that can be seen.

Yes. The closeout lists scheduled, showed, no-show, and no record against billed, not billed, waived, and late fee, with no-show dispositions and a Finalized by name. Export a day or a Monday-to-Friday week as PDF or CSV without retyping a figure. Planned, not shipped: a participant-level date-range export and GUIDE documentation at confirmation.

The preview is complimentary; we quote the rest on the call, not on a web page. About 79% of participants had some or all services paid by Medicaid in 2022 (NCHS Data Brief No. 502, 2024). The GUIDE respite allowance is $2,625 for the 2026 performance year, paid through a participating GUIDE program (CMS GUIDE Payment Methodology Paper v3.0, effective June 1, 2026). Older Americans Act dollars reach you through your Area Agency on Aging.

No savings figure. No claim that attending postpones a move to residential care; the studies we have do not support that claim, and we cite them on our research page. No claim that skilled observation stops a hospital stay; it creates chances to notice and act. Respite is not metered here; the participating GUIDE program holds the allowance.

No. Reports we prepare for funders are de-identified and aggregate; you are not given participant records. Access is scoped to each organization and to each person's own community, so you get oversight without becoming the operator.

Under our terms, the organization that brings content keeps ownership of it; we hold a license to host and display it as needed to run the service. We answer in writing, not on a web page, what leaves with a program if we stop or are acquired. We recommend asking us for that answer before the award is signed and writing it in.

Ask us for the number in writing before the award is signed, and write it in. The preview is complimentary; the price after that depends on the program, and we quote it on a call rather than on a web page. A grant should not fund software a program cannot keep.

Yes. We recommend pausing a rollout if the software adds duplicate after-hours work for staff, and we build the reports so that can be seen. Documentation gaps are measured by site and shift, never by a named individual.

Not without a dated, named source. Third-party figures carry their year and primary source, and pilot results are labelled pilot. We do not publish savings claims or outcome percentages we cannot hand you a study for. Measures are labelled by how far you can trust them: counted for you, measured with the record, or measured locally with no claim from us.

You may name Henry Ford Health System, the University of Michigan, and the Helm pilot as they are described on our studies and company pages. Cite nothing beyond what those pages say. The family daily summaries were shaped in one pilot, so we say pilot, not centers. The Day Center Companion research page carries the white paper, the claim ladder that grades each claim by how far you can trust it, and every dated figure.

Yes, as a design lens, not as clinical proof. carePhysics (our method: ten evidence-based principles about how people stay engaged) draws on named behavior-change models with a reviewer bibliography behind them. Two ways to fund: place the platform at a program you already fund, or co-apply with us.

No. We welcome companies working on connected home devices, monitors, transportation, robotics, and other technologies that support everyday care. Start with the need your product addresses and the people you want to serve.

We review the available interfaces, information flow, permissions, and support responsibilities with your team. A live connection needs a supported, configured integration; it is not assumed from the device category.

Bring a product description, your intended audience, and one practical use case. Together we can define the care experience, integration needs, and a focused pilot.

genusConnect runs on AWS under our Business Associate Agreement with AWS. Protected health information is encrypted in transit and at rest, access is logged with identifiers redacted, staff sign in with multi-factor authentication, and backups are kept. Built for HIPAA-regulated care means exactly that list, plus a written control inventory we share with your team.

We are built for HIPAA-regulated care, and we say what that means. We sign a Business Associate Agreement with AWS and with partners who handle protected health information. We encrypt in transit and at rest, scope access, and log it. On SOC 2 Type II we are at readiness: the architecture supports the criteria, and there is no report to hand you yet.

Access is scoped to your organization, to each site, and to each person's own community, and roles see their own people. We walk your IT team through exactly how during onboarding. Role names and permission detail are in the security overview rather than on this page.

AI drafts, people decide. Our AI helps your team write, find, and prepare: a draft lesson, a suggested reply, a proposed reservation change, an evening note. When AI drafts something on your behalf, such as an evening note or a lesson, a person on your team approves it first. It cannot add a fact about a person that is not in the record.

It does not give medical advice. Where an assistant answers general questions directly, it says so plainly. What the model writes passes a strict output check before anything is shown. Content the AI reads is treated as evidence, never as instructions, and any action the AI proposes waits for a person to approve it.

Sensitive exchanges and real-time analysis run on models we train and host ourselves; general tasks use leading frontier models. Identifiers are removed or minimized in prompts wherever feasible, and authors are asked not to put member data into content prompts. Details of our model providers and their agreements are in the security overview.

No. Public pages carry no protected health information and need no login; anyone can read them. Anything personal lives behind sign-in, scoped to your organization and to each person's own community.