Clinical AI · EU MDR · Portugal

Detecting deterioration before it becomes a crisis

WAVAmed builds AI-powered software for early detection of clinical deterioration in general ward and post-discharge patients, developed as a regulated medical device under the EU Medical Device Regulation.

Get in touch

The general ward is a monitoring dead zone

Admission
General Ward
ICU Transfer
Unmonitored window
Periodic checks
WAVAmed detects here
Too late

Every year, 11 million people die from sepsis. It accounts for 1 in 5 deaths globally. The majority are preventable.

The failure is not in ICU treatment. The failure is in what happens before the ICU. In general wards, deteriorating patients are not continuously tracked. Warning signs are missed. By the time standard protocols raise an alarm, the window for early intervention has closed.

Existing clinical AI tools are designed for patients already in intensive care. No CE-marked AI covers the full continuum, from ward admission through 30 days post-discharge. That is the gap where patients are lost. That is what WAVAmed is built for.

0million
Lives lost to sepsis every year. The leading cause of preventable in-hospital mortality worldwide.
1 in 0
Global deaths are sepsis-related. Most deterioration begins in the ward, hours before it becomes critical.
0hours
The early detection window we target. Identifying risk before standard protocols would raise an alarm.

What we build

Built for the ward. And beyond.

Ward-Native Intelligence
Continuous risk scoring from routinely collected clinical data. Designed for intermittent ward monitoring, not adapted from ICU tools. Post-discharge monitoring for the first 30 days, when 1 in 5 sepsis survivors are readmitted. The model flags. The clinician decides.
Regulatory-Grade Security
Deployed within hospital infrastructure. Patient data never leaves the hospital network. Built for regulatory compliance from day one, under EU MDR and GDPR governance.
Equity by Design
Model performance audited by sex, age, and clinical subgroup from the start. Closing the documented diagnostic gap for women and oncology patients, populations where sepsis presents differently and existing models consistently underperform. A design principle, not an afterthought.

Clinical evidence

Designed to satisfy regulators and clinicians.

Development Data
550,000+ patient encounters
Model developed on MIMIC-IV and eICU, two of the largest critical care databases in existence, spanning multiple US hospital systems.
Prospective Validation
3,500-bed academic medical center
Prospective validation site. Oncology general ward. Sepsis-3 adjudication committee. Performance benchmarked against NEWS2 and qSOFA.
"

The general ward is a monitoring dead zone. So is the first month after discharge. Patients deteriorate unnoticed in both. We are building the intelligence layer for both.

Hassan Seddiki  /  Founder & CEO

Let's build something that saves lives.

We are open to conversations with clinical partners, hospital systems, investors, and regulatory collaborators across Europe and MENA.

Email
contact@wavamed.com
Based in
Portugal
Incubation
Startup Braga