Your investor wants FDA.
Your US distributor is already signed.

This is the first question we resolve — before writing anything. A 510(k) requires a legally marketed predicate with substantially equivalent intended use and technological characteristics. De Novo is for Class II devices with no direct predicate — novel but lower-risk. PMA is for Class III, requiring clinical data, much longer review, and substantially higher cost. We run a classification analysis at the start of every project so you do not build a 510(k) for a device that should have been De Novo.

FDA targets 90 days of substantive review from acceptance — but AI requests pause the clock, and each round adds months. What we control: whether the submission is complete, well-organised, and passes RTA on the first attempt. What drives length beyond that: predicate complexity, software level of concern, biocompatibility scope, and whether a Pre-Sub meeting was used to align with FDA before filing. We prepare you for the AI request scenario before it happens.

Predicate selection is the highest-leverage decision in the entire 510(k). We search FDA's database by product code, intended use, and clearance date — then read the predicate's 510(k) Summary to understand what FDA already accepted as sufficient evidence for that device type. When there is a technology gap between your device and the predicate, we demonstrate the difference does not raise new safety questions and back it with performance data. That argumentation requires direct experience with FDA's Substantial Equivalence standards — not a database search.

An AI request is not a failure — it is a clarification question from an FDA reviewer, and it happens to strong submissions as well. What matters is the quality and completeness of the response. When FDA pauses the review clock, we draft the response: each answer fully sourced, documented, and with all required evidence attached. A partial response triggers another AI round and restarts the delay. We do not send partial responses.

FDA's QMSR (published 2024, effective February 2026) aligns 21 CFR Part 820 with ISO 13485:2016 — so the structural gap is narrower than it was. The remaining differences are specific: FDA Design Controls require a Design History File with a defined structure, design validation must be conducted on production-equivalent devices (not just prototypes), and CAPA expectations for FDA inspectors have particular nuances. We run a gap analysis against FDA QMSR and give you a concrete list of what needs to be added or restructured.

Yes. FDA accepts clinical data from studies conducted outside the US — including from Israeli research hospitals — provided they meet Good Clinical Practice (GCP) standards and were approved by an IRB. The critical factor is how that data is presented in the Clinical Evidence section of the 510(k). An FDA reviewer expects specific formatting and framing that reflects understanding of FDA's evidentiary standards. We specialise in translating Israeli clinical evidence into a format FDA reads as credible.

Yes — and it is better to plan them together. The technical documentation overlaps more than it appears: risk management per ISO 14971, clinical evidence, device description — all serve both submissions with adaptations in emphasis and format. We build an integrated strategy that aligns both regulatory timelines and reduces duplicate work, rather than treating FDA and MDR as two separate projects with two separate documentation sets.

FDA user fee: approximately $21,000 for standard submissions ($10,500 for qualifying small businesses). Consulting engagement: variable based on device complexity, current QMS state, software involvement, and whether clinical data is already available. In a 20-minute call we can give you a precise scope estimate. We work on fixed-scope proposals — one price, all deliverables listed. No open billing hours, no retrospective additions.