You are paying $91.23 an hour for a phone call.
OlliDirect posts your open shifts to verified clinicians directly, and leaves a record your compliance officer can actually query.
- ICU · RN07:00–19:00 · $61.00/hropen
- ER · NP19:00–07:00 · $68.00/hropen
- Peds · RN07:00–15:00 · $58.00/hropen
- Med-Surg · LPN15:00–23:00 · $44.00/hropen
Agency cost avoided
$0.00
Against the NSI 2026 average bill rate. Illustrative.
The arithmetic
The markup buys nobody anything.
The clinician does not see it. You do not get a better match for it. It is the cost of there being no direct channel between the two of you.
We are not claiming to have invented direct connection — several funded platforms already do it. We are claiming to be the one that writes down why each decision went the way it did.
What a hospital pays an agency
$91.23/hr
NSI 2026 average bill rate · 527 hospitals surveyed
Of which, the markup
30–400%
Charged over the cost of reaching the same clinician directly. The clinician does not see it, and the hospital does not get a better match for it.
Per agency nurse replaced
$66,081
Annual saving reported when an agency nurse is replaced with a direct hire — before any platform is involved.
NSI 2026 · 527 hospitals surveyed
The control stack
Five things that cannot be switched off.
The guarantees are fixed. The routing and the people are configurable — you decide who approves what, never whether approval happens.
Default deny
Every access decision runs through a policy table. Nothing is permitted because someone forgot to forbid it.
Maker–checker on money and PHI
A named human approves every payout and every release of patient data. Hard-blocked at the platform level; you configure who approves and how it escalates, not whether.
The agreement gates the data
No PHI moves for a tenant until their business associate agreement is executed, with a deliberate human confirmation before the gate opens.
Break-glass that closes again
You designate which of your roles can trigger it. Typed justification required, always time-boxed, high-severity audit entry, real-time alert to your compliance contact, never across a tenant boundary.
Isolation that does not rest on one control
Row-level security at the database and tenant validation in the application. Both would have to fail simultaneously.
The trail
Using a record and proving one are different jobs.
A live dashboard to query it, cryptographic verification to prove it has not changed, and export into your own tooling. Most platforms in this category offer one of the three.
The export itself is written to the trail.
Read the governance architecture →$ OlliDoc.it audit --shift=8842 --scope=tenant