modiolus labs
In development. Not available for clinical use.

Agentic consultation support for community ENT services

We are building regulated software that helps trained clinicians assess and manage ENT patients closer to home, under specialist supervision.

What we are building Get in touch

The problem

The constraint is not clinical knowledge. It is the number of practitioners.

ENT is among the most capacity-constrained outpatient specialties in the NHS, and most of that demand does not require an operation. Delivering specialist-grade assessment outside hospital depends on practitioners who are expensive and slow to train.

The NHS 10 Year Plan sets out a shift of care into neighbourhood settings, enabled by digital tools rather than headcount alone. The infrastructure to support that shift does not yet exist for ENT. That is the gap we work on.

75–80%

of new ENT referrals are for conditions that do not require surgery.

GIRFT, Transforming ENT Outpatient Services, 2023
~1.2 million

ENT outpatient referrals in England each year.

Figure used in our NIHR i4i outline application, 2026

The product

AURIS, a clinical decision support system for community ENT consultations.

In development as Software as a Medical Device, for use by trained clinicians under specialist supervision.

Structured assessment

AURIS takes a structured clinical history and the examination findings recorded by a trained assistant, then surfaces a ranked differential diagnosis and a care pathway recommendation.

Reasoning you can inspect

Every suggestion is presented with the evidence behind it, so the clinician can accept, modify or reject it on the same information the system used.

Clinician authorisation

Outputs remain provisional until a supervising clinician authorises them. This requirement is enforced in the architecture of the system.

A trained clinical assistant examines a patient's ear while the AURIS consultation interface runs on a screen beside them.
Illustration of the intended use: a trained clinical assistant carries out the examination at a community node while the system conducts the structured history alongside. This is an illustration, not a photograph of a deployed system.

The clinical content is authored and attested by practising ENT clinicians and anchored to published national guidance. Recommendations carry provenance back to the guidance they rest on.

AURIS is not a referral triage tool and is not intended for emergency presentations or acute airway management.

Deployment model

One supervising clinician, several community assessment points.

A trained assistant works with AURIS at each community node. A supervising GP with Extended Role in ENT authorises the outputs from a single hub.
Five community nodes, each a clinical assistant working with AURIS in a consulting room, returning provisional outputs to one supervising clinician at a central hub for authorisation.
Provisional AI-assisted outputs are returned to the supervising clinician for authorisation. The safe ratio of supervised nodes to one supervising clinician is a design assumption that has not yet been validated, and is the subject of a planned study.

Regulatory position

Classification
Anticipated Class IIb Software as a Medical Device under the UK Medical Devices Regulations. Conformity assessment not yet undertaken.
Evidence standard
NICE Evidence Standards Framework, Tier C.
Quality and safety
ISO 13485, ISO 14971, IEC 62304 and IEC 62366-1.
NHS assurance
DCB 0129 and DCB 0160 clinical safety standards, and DTAC.

Clinical evaluation is planned as a concordance study against ENT consultant assessment, reported to DECIDE-AI and STARD-AI conventions, with patient and public involvement reported to GRIPP2.

Where we are

Pre-validation. We publish the gates rather than the ambition.

AURIS has no clinical performance data yet. The sequence below is what has to be true before any claim of diagnostic performance can be made.
  1. 01Prototype and evaluation instrumentWorking prototype with a pre-specified analysis instrument, fixed before any result is generated. Complete.
  2. 02Clinical content loaded and attestedFirst condition set authored against national guidance and signed off by practising ENT clinicians. In progress.
  3. 03Retrospective vignette validationConcordance against clinician reference ratings on de-identified NHS records, inside an NHS trusted research environment. Not started.
  4. 04Prospective clinical evaluationSilent-mode concordance study under regulatory notification, supporting conformity assessment. Not started.

Team

Founded by two practising NHS clinicians.

The clinical content, the deployment model and the evidence strategy come from people who run the service the software is built for.
Dr Eason Sivayoham

Dr Eason Sivayoham

Co-founder · Clinical strategy and regulatory

GP with Extended Role in ENT and national clinical lead for community ENT. Clinical lead for an NHS community ENT service and for an independent sector ENT service.

Dr Lucy Aczel

Dr Lucy Aczel

Co-founder · Clinical content and patient experience

NHS general practice partner with a broad community caseload. She holds a diploma in dermatology and is a British Menopause Society recognised menopause specialist, extended-role community work of the kind AURIS is designed to support.

We are talking to investors, NHS partners and medtech collaborators.

If you work in regulated health software, community ENT delivery or early-stage healthtech investment, we would be glad to hear from you.

hello@modioluslabs.com