Pre-Seed · Q2 2026

Predicting surgical complications before they happen.

AI-powered real-time risk prediction for operating rooms and ICUs. Software layer on your existing monitors — no new hardware.

The Problem

What happens in operating rooms every day

An anaesthetist monitors 15+ physiological parameters simultaneously. Heart rate drifts. Blood pressure softens. SpO₂ ticks down. Each signal alone — unremarkable. Together, they form a pattern that a tired human brain, four hours into a case, may not catch in time.

~5%
of surgeries result in serious complications
higher odds of death with perioperative organ injury
+11.2
extra hospital days per complication event
Our Solution

Early warning system for the surgical journey

These risks are interconnected — hypotension drives AKI, hypoxaemia escalates into respiratory failure. A unified system tracking the full patient journey catches cascading risks that siloed monitors miss.

HIGH RISK ALERT
AI/ML
Hypoxaemia
90% probability within 10 minutes
SpO₂ trend (−3% over 8 min)
38%
Anaesthetic depth (elevated)
24%
MAP < 65 mmHg (sustained)
19%
BMI > 35 + prone position
12%

Probable causes — verify in order

Airway obstruction41%
Hypoventilation28%
Hemodynamic V/Q mismatch19%
Restrictive atelectasis12%
  • Sits on existing hospital monitors via HL7/FHIR — no new hardware required
  • Interpretable alerts with ranked contributing factors and probable causes
  • Actionable “hints” for the anaesthesiologist — reducing uncertainty and cognitive load
  • Covers intra-op and post-op adverse events in one unified system
  • No black boxes — the clinician sees why, not just what
  • Models fine-tuned to each hospital’s patient population and protocols
How It Works

Continuous risk scoring from your existing monitors

This isn’t replacing clinical judgment — it’s giving clinicians time they don’t currently have.

1

Pre-operative data

Patient history, labs, comorbidities, ASA classification, procedure type — ingested from your EHR via HL7/FHIR.

2

Real-time vital signs

BP, ECG, SpO₂, EtCO₂, temperature, ventilation parameters — streamed continuously from existing monitors.

3

AI/ML risk engine

Ensemble models trained on 10,000+ surgical cases generate rolling probability estimates for key adverse events.

4

Interpretable alert

When risk exceeds clinical thresholds: alert with ranked contributing factors. The clinician sees why, not just what.

What We Predict

Every hour of earlier intervention changes the survival curve

Adverse Event Prediction Window Status Clinical Significance
Intra-op Hypotension 10 min ahead Live MAP <65 mmHg linked to myocardial injury, AKI, stroke
Intra-op Hypoxaemia 10 min ahead Live Desaturation cascade; reintubation risk; prolonged ventilation
Acute Kidney Injury 6 hrs ahead Live 7% incidence, 25% mortality; early fluid mgmt cuts progression
Sepsis 4–8 hrs ahead In Dev Each hour of delayed antibiotics increases mortality up to 7%
Respiratory Failure Hours ahead In Dev Unplanned reintubation: 6–10× mortality increase

Current models: >90% accuracy on VitalDB and MIMIC-IV datasets.

Why We Win

Public datasets are the starting line, not the moat

Defensible

Local hospital fine-tuning

Every deployment trains on site-specific data — surgical mix, patient population, local protocols. Accuracy improves with each hospital, creating a compounding data advantage no competitor can replicate.

Novel

Capnography curve analysis

We analyse real-time capnography waveforms captured in-hospital — rich data not in any public dataset. No one else is using these curves for predictive modelling.

Unique

Full surgical journey coverage

From induction in OR to ICU discharge. Every competitor addresses a single parameter or a single phase. We cover intra-op and post-op adverse events in one unified system.

Network

World-class advisory board

Oxford Anaesthesia, Stanford AI/Anesthesia, Imperial, UCLH — our advisors shape our models and open clinical partnerships.

Clinical Traction

Building the evidence for clearance

Partner clinics today — customers tomorrow.

Current Clinical Partnerships

  • University College London Hospitals
  • Oxford University Hospitals
  • Liverpool University Hospitals
Signed MoU

Observational Phase 1 pilot study to begin on-site. Access to intra-op data including capnography waveforms.

Target: 3-6 hospital sites for multi-site clinical validation study.

Regulatory & Expansion

Regulatory pathway. FDA 510(k) for Class II CDS (predicate: Edwards HPI). Parallel MHRA Digital Health Technology route. Target market entry Q2–Q3 2028.

Global expansion. Focus markets: United Kingdom and United States. Planned extension to Nigeria and South Africa via advisors and partners.

New Module

Structured Operation Summary

Automated, structured post-operative reports from anaesthesia monitoring data. Every surgery, fully documented — review and sign off, no manual writing.

Modular — your hospital selects what to include
  • Haemodynamic overview

    Hypotension time, episodes, vasopressor log

  • Respiratory events

    Desaturations, ventilation changes, airway events

  • Fluid & blood management

    Crystalloid/colloid, EBL, transfusion events

  • Drug administration timeline

    All agents with doses, routes, timestamps

  • Case metadata

    Duration, anaesthesia time, ASA, complexity

  • Auto-generated clinical note

    Prose summary — review & sign off, not write

>10–20 min
saved per case on post-op documentation
100%
of intraoperative events captured

Clinical Audit & NSQIP

Structured data for quality reporting

Clinical Coding

Accurate OPCS-4 / ICD-10 coding

Audit-ready dataset

Queryable data for training & research

OR Utilisation

Case data for analytics & costing

Every flagged event links to raw monitoring data — clinicians can independently verify.
Who’s Behind This

Founding Team

Nikita Egorov

CEO / Science Lead

2 yrs in fintech, launched 4× offices
MSc Pharmacology, University of Oxford

Mikhail Medvedev

CTO / Engineering Lead

MSc Biomedical Engineering, TUM
BSc Biological & Medical Physics, MIPT

Dr Kristina Kleinova

CMO / Medical Lead

NHS Doctor in Neurosurgery (practicing)
MRes Surgical Sciences, University of Oxford

Advisory Board

Pavel Guzhikov

1× unicorn co-founder (Uzum)
4× product founder

Ivan Istomin

Co-Founder & COO, Roumai Medical

Scientific Advisors

Bruce Biccard

Professor of Anaesthesia, University of Oxford

Nima Aghaeepour

Professor of Anesthesia & Data Science, Stanford University

Vadim Sizov

Anaesthetist, University College London Hospitals

Hutan Ashrafian

Clinical Scientist and Surgeon, Imperial College Healthcare

Andrew Farmery

Professor of Anaesthetics, University of Oxford

David Clifton

Chair of Clinical Machine Learning, University of Oxford

Jacob McKnight

Health Systems researcher, University of Oxford

François Taljard

Anaesthetist NHS, CEO & Director of Multiple Companies

Every surgical team has a moment where they wish they’d known
five minutes sooner.
We’re building the system that gives them that time.
Contact Us

Let’s talk

Interested in partnering, investing, or learning more? Reach out to our team.

Nikita Egorov

CEO & Co-Founder
nikita.egorov@st-hughs.ox.ac.uk

Mikhail Medvedev

CTO & Co-Founder
mikhail.medvedev@tum.de

Kristina Kleinova

CMO & Co-Founder
kristina.kleinova@st-hughs.ox.ac.uk