What operators ask us.

With a conversation, not a demo reel. Write to [email protected] or use the contact form, and a person answers within 24 hours. The first call is about where the work breaks down — the shifts you can’t fill, the agency invoices, the denials, the survey findings.

If it’s a fit, the next step is a working session against your own numbers, and from there a scoped rollout with success criteria agreed in writing before anything goes live.

Start a conversation →

Hospital systems, post-acute operators, and rural and critical-access facilities. The engine is the same; the budget line differs. Hospitals buy capacity protected — staffing flexed to the census you’re about to have, handoffs closed on time. Post-acute operators buy margin protected — gaps filled from your own bench at W-2 rates instead of agency premiums. Rural operators take the whole thing delivered as a service, because the smallest teams have no one to spare to run a transformation.

If staffing gaps on your floors turn into agency premiums, documentation gaps, and denials, you’re who we built it for.

No. Remedi serves a single critical-access hospital and multi-facility chains alike, and the engagement is sized accordingly. Larger systems usually bring their own IT and security review; smaller operators take Remedi fully hosted and fully run. Neither posture is second-class — same product, same governance, same support.

Sixty to ninety minutes with the people who own the problem — typically the COO or CFO, the DON or CNO, and whoever runs your systems. We walk through your numbers, not our slides: your agency spend, your overtime, your fill times, your denial mix.

You see the product working on realistic demonstration data — never your PHI before agreements are in place — and you leave with a dollar model built on your own figures.

Weeks, not quarters. The first connection is read-only and typically live within a few weeks of signed agreements — Remedi forecasting and proposing alongside your existing process while your team compares. From there, go-live proceeds unit by unit on your schedule. A full proof period — baseline, success criteria, weekly operating reviews — runs sixty to ninety days.

A scoped engagement with the ending written first. Named facilities and units. A baseline pulled from your own history — agency spend, overtime, time-to-fill, denial rate — agreed before go-live. Success criteria in writing, weekly operating reviews against them, and clean exit terms: if the numbers don’t prove, you keep the baseline analysis and your data, and we part as professionals.

Three things you can verify, not take on faith. The outcomes we’re accountable for are written into the contract, not implied by a slide. The system shows its work — every action logged, overridable, and cited to a rule and a model version — so you never have to trust it blind. And the proof is made on your floor, against your own baseline, not on our references.

Three things: confirm your facilities, name the systems you use, and approve access. Remedi’s team does the rest — configures and tests every connection, validates the data flowing through it, and seeds the first schedules.

Your IT sees a checklist measured in hours, not a project plan measured in quarters.

Zero — and that’s engineered, not promised. No sandbox to provision. No PHI-scrubbing or de-identification project. No configuration workshops, no data-mapping spreadsheets. No models to train. No infrastructure to stand up, patch, or babysit. Remedi’s team carries every technical motion, white glove, from first connection to steady state.

The effort you spend is the kind only you can spend: organizational change and workflow decisions — what should run differently, who approves what, where the thresholds sit — with the impact measured in the real world, against your own baseline.

Yes — that’s the posture rural and critical-access operators take by design. Remedi is delivered as a service: hosted, configured, connected, monitored, and tuned by us. What we need from you is someone with the authority to approve access, and a leader on the floor who will tell us the truth about what’s working.

The same three people who already run everything else in your building don’t take on a fourth job.

No. There is no migration. Remedi reasons over your systems of record where they already live — it reads the EHR, the schedule and time systems, and the clearinghouse, and writes back through the same connections. Nothing is exported, re-keyed, or cut over. Your history stays exactly where it is.

By category: the EHR and the AI scribes your clinicians dictate into; the scheduling, time, and workforce systems your staff clock into; your clearinghouse, for eligibility, claims, status, and remittance; the nationwide clinical record network; and your payment gateway. Connections speak FHIR, HL7, X12, and direct APIs — real-time or batch.

If a system holds your schedule, your chart, or your claims, connecting to it is our job, not yours.

See the full stack →

Nurses install Remedi Go from a link on their own phone and are claiming shifts in their first session — if a screen needs a manual, we treat that as a defect. Operators get working sessions in the product, on your data, during the first weeks.

And because every recommendation carries its reason with it, the product explains itself as it works — training that never really ends, without a classroom.

Through your identity provider, with role-based access enforced server-side — a nurse’s schedule, a charge nurse’s approvals, a physician’s panel, and a billing specialist’s queue are different systems as far as the data is concerned, not different menus. Access is provisioned at go-live, follows your HR reality, and every session is logged.

Deliberately boring. It starts read-only: Remedi forecasts and proposes alongside your current process, and your team compares. Then unit by unit, facility by facility, you let it act — approvals first, autonomy where you decide it’s earned.

Your existing process keeps working until you choose to lean on the new one. There is no big-bang Tuesday, and nothing breaks.

Scheduling is one verb inside the job. Remedi manages the workforce end to end — forecasts demand shift by shift, solves a fair schedule, fills gaps from your own bench first, and controls the cost in dollars your finance team trusts — then manages the work the workforce touches: documentation completeness, eligibility, claims integrity, continuity of care.

A scheduling tool stops at the roster. Remedi is accountable for what the roster costs and causes.

No. Run as many or as few of the workflows as you want — workforce only, workforce plus eligibility, the full surface — and the same goes for the client apps. Each one stands on its own and plugs into the same connections and governance underneath.

Most operators start where the budget bleeds worst and add from there. Expanding is a decision over the ledger, not a second implementation.

Remedi Go, on their own phone: their schedule, open shifts to claim, swaps to propose, the preferences their scheduling agent respects, and their own fairness ledger — the same math their manager sees, not a black box. It installs like an app, works offline, and never asks them to learn “software.”

Adoption is designed in, not hoped for. A nurse gets something for themselves on day one — the schedule in their pocket, shifts without phone tag — and the fairness ledger is the trust move: staff who can check the books stop suspecting the house. The preferences they control feed an agent that works on their behalf, so the system reads as an advocate, not surveillance.

That’s why fairness is the first screen, not a settings page.

The operating picture, live: coverage against the demand forecast, compliance findings before they become survey findings, claims and denials traced to their causes, and the ledger of everything the agents did — each entry with its reason, its rule, and its cost. Approvals queue where judgment is required.

It reads like a well-kept ledger, because that is what it is.

Authority runs in order: deterministic math where precision matters, learned models where behavior matters, Claude reasoning where nuance matters, and your team’s governance where accountability matters. Nothing clinical or financial commits without passing its gates.

Every action is logged, overridable in one motion, and citable to a rule and a model version — and autonomy is granted by you, scope by scope, never assumed by us.

It sometimes will — we designed for that instead of around it. Hard gates keep low-confidence work from committing on its own. An override takes one action and goes on the record. And because every decision cites its rule and model version, a miss is traceable, correctable, and stays corrected.

What you should refuse to buy is AI whose mistakes can’t be found.

Never. The systems of record — your EHR, core HR and billing, the rails that carry your data — stay exactly where they are; Remedi reads and writes through them. What gets retired is the sprawl stacked around them: the standalone scheduling, eligibility, denials, prior-auth, and dashboard tools your teams keep a dozen logins for. Their work moves into Remedi, and each subscription gets cancelled.

Count the logins. The standalone scheduling portal, the eligibility checker, the denial workqueue, the prior-auth tracker, the dashboard subscription bought to watch all the others — one by one, as Remedi absorbs the work, each becomes an invoice that stops arriving.

Killed license fees post to the same ledger as avoided agency spend, so the consolidation is measured, not asserted.

No — we sit on your side of the table. Remedi never supplies labor. It fills gaps from your own bench first — full-time, part-time, PRN, float, matched by competency and distributed fairly — and escalates to the agencies and marketplaces you already contract with only when the bench is truly exhausted, under your rates, with the cost differential logged.

A marketplace profits when you use it more. Remedi proves itself by helping you use it less.

Yes. Remedi ships a Model Context Protocol (MCP) server, so the assistants your organization already uses — and the ones you build — can call Remedi’s tools directly: schedule state, fairness positions, compliance findings, claim status. Machine callers get the same treatment as human ones: role-scoped access, the same hard gates, and every call on the audit ledger.

Decades of workforce-operations research inside large hospital systems, and the published organizational-justice literature that identifies fairness — not pay — as the actual driver of nurse attrition. The citations are on this site, not in a vault.

And the science is being tested in the open: a joint research initiative with Middle Tennessee State University is studying whether schedules built by AI that explains its decisions are experienced as fairer and more trustworthy than optimization alone.

Wherever your governance requires. Most operators run fully hosted in Remedi’s cloud — built on Google Cloud, isolated tenancy, managed end to end, the full-service posture. Systems with their own cloud standards run Remedi inside their own environment, on any major public or private cloud. And for the strictest postures, Remedi deploys on-premises.

Same product, same governance, same audit trail in all three.

Remedi is built Claude-native on Anthropic’s Claude, and the reasoning layer is replaceable by design. If your organization has standardized on a different frontier model — or requires a private model hosted inside your own boundary — Remedi works with it. The deterministic rules, the fairness math, and the governance don’t change with the model, and every narrative still cites its sources or stays silent.

The model is a component. The discipline is the product.

Yes, and yes — a business associate agreement comes with every engagement, and verifying that BAAs are current is literally one of our compliance rules. PHI stays under your tenant boundary, encrypted in transit and at rest, with access role-scoped, least-privilege, and logged.

The full security documentation — architecture, subprocessors, incident response, data retention — is available for your review under NDA.

No. Your data is never used to train foundation models, never pooled across customers, and never sold. Model calls run inside the same governed boundary as everything else, under the BAA chain. And what the reasoning layer produces is cite-or-omit: a narrative may only say what your record supports, with citations back to the source — anything it can’t cite, it doesn’t say.

You own it — full stop, written into the terms. Export what Remedi holds, in standard formats, whenever you like — not just at exit. If you do leave: connections shut down cleanly, your export is delivered, and destruction is attested in writing.

And because your systems of record were never moved, leaving Remedi never strands your history. We hold ourselves accountable for outcomes, not hostages.

Yes — sharing is designed in, not an export ritual. Governed Remedi data can be shared zero-copy in Snowflake or published as open Apache Iceberg tables, so your analysts and data teams query it in place, with the tools they already use — no ETL project, no second copy drifting out of date.

The same role scoping and audit trail apply on the way out as on the way in.

The record that does the work is the record you audit. Remedi executes as durable, governed workflows whose replay is the audit trail — it cannot be incomplete — and every action carries its rule and model version. When the surveyor, the auditor, or counsel asks why the system did what it did, the answer is an export, not an archaeology project.

Remedi holds its state and waits. Workflows are durable and exactly-once: a step that can’t complete resumes when the system returns — it never double-executes and never silently drops work. The floor keeps its schedule, staff keep their apps, and reconciliation happens on the record when the connection is back. Uptime and support commitments are written into the agreement.

Layered like the product. AI support agents that know your configuration answer immediately, around the clock — and anything they can’t resolve escalates to named people who know your account, not a queue with a stranger behind it. Every request lands in a shared tracker both teams can see, so nothing disappears into a portal. Response commitments by severity are written into the agreement.

It means the division of labor is explicit: your team runs the floor; we operate Remedi. Hosting, connections, monitoring, tuning, model updates, new capabilities — ours to operate, with the outcomes we’re accountable for written into the contract.

Not a system you assemble and staff. A result we stand behind.

Continuously — there is no version to buy, no upgrade project to schedule. Improvements arrive as they are proven. And because every behavior is governed, anything that changes how decisions get made shows up on the ledger with its new rule or model version — announced, visible, and reversible.

Through the same shared tracker your support lives in — you see the status of what you’ve asked for — and through the weekly and quarterly operating reviews, where priorities get argued honestly. Operators who work with us early shape the work. That’s the deal, and it’s why we call them partners rather than accounts.

An annual subscription per facility, sized to the scope it runs. Clinical staff are never charged per seat — every nurse gets Remedi Go, because fairness only works when everyone can see it. A proof period is priced on its own terms, with the success criteria attached.

No meters that punish you for using the thing.

The subscription covers the workflows you choose, the client apps, the connections behind them, and the outcome reporting that proves the ledger. Around that sits a menu of optional services — data sharing into your warehouse, the fax rail, AI-scribe capture, patient-pay collection, advanced analytics — switched on only when you want them, priced only when they’re on.

You use what you need. Nothing rides along on the invoice.

You’ll see it in your own ledger, not our slide deck. The baseline is agreed before go-live — agency spend, overtime, time-to-fill, fill-from-bench rate, turnover, denial rate, and the subscriptions you pay around the record — and Remedi reports against it continuously, in dollars your finance team already trusts.

We deliberately publish no outcome claims we haven’t earned. The proof period exists so your numbers can do the talking.

The market numbers say where the money is: agency and travel labor bill 1.5–2.5× your own staff; 17.6% of registered nurses turn over every year at roughly $60K a replacement; a fifth to a quarter of Medicare Advantage post-acute claims are initially denied. Those are the levers Remedi pulls — bench-first fill, fairness that slows the turnover, denials traced to their causes.

How far they move on your floors depends on your mix — and measuring that honestly is precisely what the proof period is for.