ISelf-declared path
Class ICore technical file structure first.
Organize Annex II structure, labeling, IFU, and evidence readiness before you enter a more formal review cycle.
Typical focus: file structure, labeling control, intended use clarity.
RunaReg maps your technical file to MDR 2017/745 and latest EU requirements, surfaces compliance gaps, and drafts evidence-backed remediation with deterministic rules.
Explore our live, step-by-step interactive product tour covering clause-linked gap review, PRRC signatures, and pre-submission audit readiness.
Click play on the simulator to walk through real workspace views, database checklists, and regulatory graph trails built to secure MDR compliance.
An interactive visual walkthrough showing the transition from regulatory bottleneck chaos to clean OS compliance.
RunaReg supports the full MDR pathway, with particular depth where teams lose time: special Class I controls, Annex VIII logic, and higher-risk preparation before a Notified Body review starts.
Organize Annex II structure, labeling, IFU, and evidence readiness before you enter a more formal review cycle.
Keep GSPR mapping, risk documentation, and evidence organization ready for early notified-body-facing review.
Track stronger clinical and risk evidence, gap remediation, and reviewer-ready documentation before the formal clock starts.
Prioritize CER, PMCF, PSUR, and clause-level traceability so the file is defensible before higher-intensity review begins.
Instead of treating Class I as one bucket, RunaReg surfaces the extra controls for sterile, measuring, and reusable surgical devices before they become review delays.
RunaReg frames software by intended use, clinical impact, and documentation burden, not by “software” as a shortcut label.
Three checkpoints from regulation text to a reviewable finding. The same input always produces the same clause link.
MDR 2017/745, Annexes II / III / VIII / XIV, MDCG guidance, harmonized standards — versioned in our regulatory knowledge graph.
Each clause is linked to the matching paragraph in your tech file. Every match is reproducible and the trace is exposed for review.
PRRC or consultant approves remediation. Nothing exits the platform without a human regulatory owner.
RegOS applies deterministic clause mapping first. Draft remediation remains review-only and never replaces a qualified regulatory owner.
The determinism comparison below shows what changes when your output has to be reproducible at audit — not just persuasive on screen.
Three steps from upload to NB-ready, each linked to a clause and a reviewer signature.
Upload your tech file (PDF or DOCX up to 50 MB). RunaReg classifies the device, structures the file against Annex II, and surfaces what is missing.
Every paragraph is checked against the 200+ MDR sub-requirements a Notified Body assesses. Each gap is clause-linked and severity-graded.
Once you submit, the EU 2026/977 clock budgets (30 / 120 / 90 / 20 days) are tracked in-app, so you know exactly when a finding will pause the clock.
The digital center of your MDR journey. Built for the two people who actually own MDR readiness: the manufacturer's PRRC, and the external consultant they hire. Same evidence, same clause links, same export.
Upload your technical file. RunaReg surfaces every Annex II gap, drafts remediation text for review, and tracks your NB timeline so your PRRC always knows what to fix first.
RunaReg does the deterministic mapping. You provide the strategic judgment. Findings arrive clause-linked, exportable, and pre-formatted for client workshops — with your name on the cover.
Consultants pay a flat monthly platform fee and can invite unlimited clients.
Manufacturers pay only for active device families. No per-user fees.
Enterprise supports single-tenant deployment in your own cloud.
Ten yes/no questions on your device, documents, and process. We score the signal against the same readiness model used in the platform, then send the detailed report to your email.
Limited self-check. For a full clause-linked score on your actual documents, use the freemium analysis lane or request a demo.
RunaReg’s Knowledge Graph encodes MDR, IVDR, UKCA, and Turkey’s national regulation. Select your target markets during device registration, and the platform automatically surfaces the right requirements—not a generic checklist.
Class I to III pathways, Annex II/III structures, and Article 120 transition tracking.
Knowledge Graph EncodedIn vitro diagnostic classifications, Annex II structures, and performance evaluation files.
Knowledge Graph EncodedUK Medical Devices Regulations 2002 alignment, designated standards, and UK Responsible Person gates.
Knowledge Graph EncodedTurkish national legislation alignment, TITCK product registration, and 2026 enforcement deadlines.
Knowledge Graph EncodedOur hybrid pipeline combines graph-guided query expansion, dense semantic search, and exact clause matching via BM25—fused with Reciprocal Rank Fusion. The result: every document draft is grounded in the exact MDR clause, MDCG guidance, and harmonised standard that applies.
Gap finding trigger
2-hop walk of GOVERNED_BY & CLARIFIES
Similarity on MDCG guidance PDF chunks
Exact clause & standard match
Score fusion & deduplication
Contextually validated CER/GSPR text
Notified Bodies reject submissions when contraindications differ between the IFU and the Clinical Evaluation Report. The Consistency Agent detects these mismatches across your entire Technical File before submission, so you fix them first.
Marked N/A — Claims 'Information supplied' is not applicable.
IFU document present — An actual IFU file exists for the device family.
GSPR_IFU_MISMATCH: GSPR clause 23 is marked N/A in the checklist, but an IFU document is present. This is a direct contradiction. Auditors flag this as a major non-conformity.
Marked Applicable — Refers directly to IFU document version 1.4.
IFU document present — No contradiction detected. Document trace is complete.
The Drafter generates text with inline structured assertions. A schema-bounded validator checks each one against the Knowledge Graph. If a claim isn’t supported, the system retries with more specific regulatory context—before you ever review the draft.
The generator outputs regulatory text containing inline structured assertions linking specific claims to Knowledge Graph nodes.
The Judge checks assertions: KG node must exist in Neo4j, confidence must meet the 70% threshold, and quotes must match the source verbatim.
Unverified assertions generate machine-remediation hints for auto-retry. Verified text is locked, signed off by the PRRC, and stored.
We'll publish numbers when the pricing model is fully validated with the early cohort. The shape is fixed.
Flat platform fee with multi-client workspaces.
Co-edit with your manufacturer's PRRC. Branded report exports for client handoff.
Tied to active device families in the lifecycle.
You only pay for product lines under active NB preparation or post-market work.
Deployable in your own EU secure cloud.
Single-tenant deployment, SSO, dedicated environment, contract-level SLAs.
Anonymized sample. Same structure your team and consultants carry into PRRC review and NB preparation.
Reconcile shelf-life and storage claims across IFU, label, and risk documentation.
Top-line signal showing severity concentration and where the package needs work first.
Each gap ties back to the governing MDR clause. The reviewer never has to reconstruct the logic.
Suggested next step in a form a PRRC can review, edit, and adopt directly.
Device identity, company name, and identifying details are removed.
Quotes from the early-access cohort. Names available on request — we are deliberately conservative about public attribution while teams are still onboarding.
The clause-linked output changed the conversation from guessing to evidence. We could triage Annex II gaps before the NB ever saw the package.
What mattered was not speed alone, but traceability. Our PRRC could inspect why each issue was raised instead of trusting a black box.
The difference versus generic AI was the structure. Findings, remediation, and evidence references were already in a reviewable format for client workshops.
From device classification through post-market surveillance, RunaReg modules connect into a single operating system — each module owns a defined lane in your file.
Classification & structural framing
Evidence review & document drafting
Lifecycle, remediation & surveillance
No. RunaReg structures the evidence, findings, and remediation workflow. A qualified human reviewer remains the regulatory owner and decision-maker.
The same file and the same rule set produce the same clause link. The platform does not guess — it applies predefined matching logic and exposes the trace for review.
Yes. The platform is designed for pre-submission readiness, gap identification, and remediation planning before a Notified Body starts the formal clock.
Frankfurt-based EU hosting with GDPR-conscious operational posture. Customer documents are isolated per workspace; audit log is ALCOA+ oriented.
Generic AI tools answer from training-time corpora. RunaReg answers from a versioned regulatory knowledge graph — MDR, MDCG guidance, and harmonized standards represented as nodes and clause links, rebuilt on every regulation update. When the platform retrieves an answer, it retrieves from the graph (a Graph-RAG architecture), then checks the LLM's draft sentence-by-sentence: does the clause exist as a node? Does the cited standard? Is the evidence in your file? Same input, same graph version, same answer. That's what makes it audit-defensible.
You receive a scored signal by email. If you want the full clause-linked score on your actual documents, we coordinate a demo and onboarding.
Tell us your device class, submission stage, and current bottleneck. We will reply directly to coordinate access — no automated drip, no demo theatre.