Preference cards drift from reality
Cards are written once and used for years. Kits arrive with items nobody opens and without the one the surgeon asks for.
For ambulatory surgical centers
The autonomous procurement system for ambulatory surgical centers. AI agents read the surgical schedule, match every item to your contracts, check its regulatory status and place the order.
Westbrook Surgery Center
Tomorrow, 18 cases across 3 rooms
426
98.6%
Tue, 07:30
4 orders in transit, all on time
Surgery has robots in the operating room. Its supply chain still runs on faxes, sticky notes and sales reps walking the halls.
6,100+
ambulatory surgical centers in the United States
28M
procedures they perform every year
4–6h
of staff time behind a single manual order
Cards are written once and used for years. Kits arrive with items nobody opens and without the one the surgeon asks for.
Negotiated pricing lives in one place and purchasing happens in another, so off-contract spend slips through.
Shortages and recalls are discovered at the case cart, when the only options are a rush order or a delay.
mintHealth is a network of specialised AI agents. Each one owns a single job, and none of them waits for someone to update a spreadsheet.
Step 1 of 4
Cases and preference cards flow in from the EHR and the scheduling system. mintHealth forecasts what each procedure will consume, case by case.
Step 2 of 4
Every line is matched against GPO and local agreements, so orders go out at the price the center has already negotiated.
Step 3 of 4
Recalls, FDA status and device identifiers are verified on each line before anything is ordered.
Step 4 of 4
Approved orders are placed automatically and tracked to delivery. Exceptions go to a person, with the context attached.
Every order passes hard-coded checks before it leaves the building. Nothing reaches the operating room on a model's best guess.
Orders for recalled or flagged devices are blocked before they are placed.
Device identifiers and lot numbers are verified before anything enters the OR.
Unexpected supplier fees, such as sterilization or freight, are surfaced next to the contract price.
Stop reconciling preference cards and re-keying orders. Handle the exceptions, not the routine.
See committed spend against contracts before it happens, per case and per surgeon.
Start each case with the right items on the cart and fewer day-of-surgery surprises.
Regulated healthcare can't learn on live patient data first. So we build the simulation, prove the logic, and only then deploy. Developed in Warsaw.
Agents learn logic, negotiation and compliance rules on statistically faithful synthetic twins of clinical and operational data, before they see a live system.
Our Adaptive Root Architecture grows or prunes pathways as data density changes, so models stay lean as they scale.
Core models carry no personal data. Each deployment is scoped with the partner's compliance team and starts under a mutual NDA.
Targets we set with each center at kickoff and report against, line by line.
95%
of spend on contract
<30 min
from booked case to placed order
<1%
order line error rate
0
recalled devices ordered
No. It takes over the repetitive work, such as reconciling preference cards, reordering and checking prices, and hands anything that needs judgment to your team with the context attached.
Your EHR and scheduling system for case data, and your purchasing channels for orders. The exact integrations are agreed during onboarding.
mintHealth uses only the minimum case data needed to forecast supplies. Its models are built on synthetic data, and every deployment is scoped with your compliance team under an NDA.
Per surgical center, based on case volume and integrations. Book a call and we will send you a quote.
A short call to understand your center, then a walkthrough with your cases. NDA available from the first conversation.