Medical Device Declaration

1. Regulatory classification

MedPro AI is CE-marked as a Class IIa Medical Device under EU Medical Devices Regulation 2017/745 (MDR), classified per MDR Annex VIII Rule 11: "Software intended to provide information which is used to take decisions with diagnosis or therapeutic purposes is classified as class IIa."

• Manufacturer: MedPro AI Ltd (trading name of DJG Media Limited), CRO 762838, Coliemore House, Coliemore Road, Dalkey, Co. Dublin, Ireland
• Device name: MedPro AI Clinical Practice Platform (including the Brigid AI assistant)
• GMDN code: 60918 (Health-care management software)
• UMDNS code: 16-262 (Medical Records System Software)
• Notified body: pending public registration — see s.6 below
• Basic UDI-DI: pending publication on EUDAMED — see s.6

2. Intended use

MedPro AI is intended for use by registered healthcare professionals to support clinical practice in primary care, dentistry, and adjacent allied-health settings. The platform provides:

• Electronic medical record (EMR) functionality — capture, retrieval, sharing
• Appointment scheduling, billing, and clinic administration
• AI-assisted clinical decision support (the "Brigid" component) — drafting of letters, summaries, referrals, and structured clinical notes
• Voice transcription of clinician-led consultations (consent-gated, EU-resident)
• Patient-facing companion app (MedYou) for record access, messaging, and consent

What MedPro AI is NOT:it is not an autonomous diagnostic device. It does not prescribe medication, sign documents, or finalise clinical decisions without a registered clinician's review and approval. AI outputs are always drafts. A qualified human must verify every output before it is acted on.

Contraindications + limitations:the platform is not validated for paediatric oncology, emergency-medicine triage, ICU monitoring, or any acute clinical context where lives depend on sub-second response. Use is limited to the conditions described in the platform documentation and to the clinician's licensed scope of practice.

3. EU AI Act overlay — high-risk classification

The Brigid AI component falls under Regulation (EU) 2024/1689 (EU AI Act) as a high-risk AI system by virtue of two overlapping classifications:

• Annex I list — Brigid is a safety component of an EU MDR medical device (the platform).
• Annex III point 5 — AI systems intended to be used by public authorities or on their behalf to evaluate the eligibility of natural persons for essential public services (health). Even where Brigid is used in private practice, the functional classification stands.

Where Annex I and Annex III overlap, we apply the stricter set of obligations of both.

4. Human oversight duty (EU AI Act Article 14)

Every AI output must be verified by a registered clinician before it is acted on. This is not optional. It is enforced as follows:

• The platform never auto-sends, auto-signs, auto-prescribes, or auto-actions an AI output. Every artefact requires explicit clinician approval.
• Clinicians accept this duty as part of the in-app Clinician Consent Gate (Medical Device & AI human oversight category) before they can use the platform.
• Misuse — attempting to bypass review, e.g. auto-routing AI drafts to patients without clinician sign-off — is grounds for immediate suspension of platform access and referral to the relevant professional regulator.
• Clinical patterns suggestive of bypass (low approval rate, near-zero edit distance between AI draft and patient-facing artefact) are monitored as part of post-market surveillance.

Patients are informed of AI use at multiple points: the first-sign-in consent gate (AI Features category), the persistent off-state banner when AI is declined, and the visible Brigid orb avatar — satisfying EU AI Act Art 50 transparency.

5. Post-market surveillance & audit log

We operate a continuous Post-Market Surveillance (PMS) system under MDR Articles 83–86 and a Post-Market Performance Follow-up (PMPF) plan per Annex XIV.

• EU AI Act Art 12 — append-only audit log of every AI inference, with input/output category, timestamp, clinician approval status, and the model version. Retained for the lifetime of the device plus 10 years.
• Adverse event reporting — incidents that meet the MDR Art 87 reporting criteria are reported to the competent authority within the regulatory deadline (initial within 2 days for serious incidents).
• Clinical Evaluation Plan + Risk Management File per ISO 14971 are maintained and reviewed annually or upon any substantive change.
• Trend report — submitted to the notified body annually summarising non-serious incidents and performance metrics.

6. Unique Device Identifier (UDI) — pending

Under MDR Article 27 we are required to assign a Basic UDI-DI and a UDI-DI to the device and publish them on EUDAMED. This is in progress alongside notified body certification. The current placeholder values are:

• Basic UDI-DI: pending — assigned by issuing entity per ISO/IEC 15459-2
• UDI-DI:pending — published on the device label and within the platform "About" screen once finalised
• Notified body: pending — to be named on this page and on the EUDAMED entry
• SRN (Single Registration Number): pending

We commit to updating this page within 30 days of any change in regulatory status. For the current status, contact our Quality Management System team at qms@medproai.com.

Related documents

Clinician Terms of Service · Data Processing Agreement · EU AI Act compliance · AI Features (patient privacy) · Acceptable Use Policy

To report a vigilance issue or adverse event: vigilance@medproai.com. For Quality Management System enquiries: qms@medproai.com.