VitVio

Everything Vetro does comes from its 3D model of the room. A live point cloud of each operating room, built from markerless calibration, lets the system follow people, equipment and activity in depth and through occlusion, so it can log surgical stages, keep the block schedule current, recognize the instruments on a tray and draft the operative note from voice and activity. Mission Control, the tray data and the op note all run on that spatial model.

The company sets its approach against two dimensional detection systems that see flat bounding boxes and only basic events such as wheels in, wheels out and draping.

Most of Vetro acts on its own. It logs surgical stages and writes them back to the record without anyone reporting them, updates the schedule as the day unfolds, tells the pre op ward when a case is closing and the next team when a room is clean, and the company describes this orchestration as autonomous. The surgeon reviews and signs the operative note before leaving the room.

The site also says every instrument is accounted for and that the note is written by a system that never loses context, with no stated error rate or check for a missed instrument or an omitted detail.

The 3D Spatial Context Engine reasons over point clouds, body poses and room geometry rather than flat images, keeps spatial memory across the room and handles occlusion continuously, while the operative note draws on voice and activity captured as the case runs. The company describes Vetro as 3D native and has a chief scientist, Grzegorz Jacenków, behind the engine.

No model, training data or version is named.

Microsoft Azure is the named subprocessor for cloud infrastructure and hosting, under data processing agreements with security, confidentiality and GDPR clauses, and the company's own 3D Spatial Context Engine does the room modeling. Other providers of hosting, analytics or technical services may receive data as each hospital authorizes.

Whether the operative note relies on an outside language model is not stated.

Outside benchmarks stand in for results from VitVio's own deployments. The turnover figure placed beside the claim that autonomous orchestration shortens turnover, 37 minutes down to 14, links to an Annals of Thoracic Surgery article rather than to a Vetro site, and the cost per minute and revenue per hour figures come from other published work. Quotes from clinicians in London, Los Angeles and Spain speak to operating room efficiency, workflow and data collection.

Vetro has no named deployment or measured result on the site.

Footage is seen once and then gone, the company says: most video is discarded moments after processing, with no archive, playback or library of OR video, and the engine works on point clouds, poses and geometry that never need faces or identities. The privacy policy, updated May 15, 2025, keeps personal information only as the hospital directs and for no longer than 90 days, and patient age and weight categories are processed on the hospital's instructions to improve predictions.

Whether customer data trains Vetro's models, and how a safety event involving the system is handled, are not stated.

The privacy policy, dated May 15, 2025, covers the product as well as the website. Hospitals are the controllers of patient, staff and video data; VitVio processes it on their instructions under a master services agreement, handles breaches under GDPR and HIPAA procedures, and has appointed a HIPAA security officer, with VitVio Ltd in London as the contact.

The policy describes a GDPR processor and never states business associate status for US hospitals.

VitVio's privacy policy describes encryption, restricted access, physical security controls and confidentiality obligations for staff who handle customer data, with breach procedures under GDPR and HIPAA. A trust center on its own subdomain opens to visitors who request access, and the home page shows four badges marked certified.

The standards behind those badges are not named in the page text.

Vetro is sold as operating room automation, covering stage logging, scheduling, tray recognition and documentation, and no device claim is made. The operative note comes closest to clinical content: it is drafted continuously from voice and activity in the room, synced with the record, and reviewed and signed by the surgeon before it is final.

How the company positions note drafting and instrument accounting under FDA or UK device rules is not stated.

Vetro is designed to reduce administrative burden and never to monitor individuals: its outputs describe cases, phases and equipment as team level patterns rather than personal performance, and each hospital sets its own retention rules and access policies. The privacy policy names a data protection officer and a HIPAA security officer.

No evaluation of how the models perform across room layouts, specialties or hospitals is published.

The surgeon reviews and signs each operative note before it is final, and the note carries its sources and codes so a reviewer can see what it was drawn from. Patients and staff send correction and deletion requests to the hospital, and VitVio assists as contract and law require.

No accuracy figure, warranty or remedy for a wrong stage, tray count or note is published.

Logged stages go back to the hospital's record system automatically, with Epic and Cerner named, and the operative note is shown synced with the EHR, its sources and codes attached for the surgeon. Stage events are shared across the perioperative floor so pre op and recovery teams see the same state as the room.

No interface standard or live customer integration is named.

Microsoft Azure hosts the platform under data processing agreements, and data may be stored outside a user's home country with contractual safeguards agreed with each hospital. Each hospital sets the retention rules and access policies, and most video is discarded moments after processing.

Which Azure regions hold data, and whether processing runs on site or in the cloud, are not stated.

Licensing runs business to business, with hospitals and health systems signing a master services agreement that sets out how operating room video, audio and logged events are used for each customer. The home page frames the return with outside figures: operating room time at about $53 a minute and more than $6,000 in revenue for each recovered hour, each linked to a published source.

No price, unit of charge or hardware cost is published.

Hospital operating rooms and the perioperative team around them are the target: surgeons, nurses, anesthesiologists, OR managers, and the pre op and recovery teams that wait on each room. Examples on the site lean toward orthopedic lists such as knee, hip, shoulder and ankle procedures, and the quoted clinicians are a London knee surgeon, a Los Angeles director of anesthesiology and a Spanish head of obstetrics and gynecology.

The company says it works with hospitals in the US and Europe without naming any of them.