Treat or Prevent
Drives therapeutic action — closed-loop insulin dosing, ventilator control.
Four intended-use categories push software into SaMD territory. Your intended use statement draws the line — not the label.
Drives therapeutic action — closed-loop insulin dosing, ventilator control.
Identifies disease or injury — AI imaging, arrhythmia and sepsis models.
Tracks clinical state for immediate action — real-time deterioration alerts.
Risk scores that drive decisions — treatment choice, escalation timing.
Six categories — each with its own regulatory pathway and clinical-validation requirements.
Most diagnostic AI and CDS is Class II.
Risk and seriousness set the class.
Predicate exists — fastest route.
No predicate — FDA sets a new class.
FDA feedback before you build.
Clinical validation is a study, not testing.
Academic-center data may not generalize to the community setting you deploy in.
Sensitivity, specificity and PPV/NPV at realistic prevalence — not raw accuracy.
FDA expects equitable performance across demographic and clinical subgroups.
Post-market real-world data weighs heavily for AI/ML products. We build for it.
Each card is a SaMD product we designed and cleared.
Talk to Our TeamFDA device law, lifecycle standards, and risk management — built in from the start.
SaMD guidance and 510(k) / De Novo / PMA pathways.
Design-control requirements an FDA submission depends on.
Lifecycle process, risk management, usability, and cybersecurity.
PHI handling, encryption, access controls, and audited security.
EU MDR Article 22 SaMD provisions plus GDPR compliance.
Clinical data exchange, imaging, and accessibility.
Organizations that discover SaMD requirements during development build them in. Those that discover them at launch redesign under pressure and delay market entry. The engagements that go well start with regulatory strategy. Thirty minutes. No pitch.
Book a Discovery Call
100 Fastest Growth Companies
Global Spring Winner
Top App Development Company
AWS Partner Network
Google Cloud Partner
Highly Rated on Trustpilot
Verified Agency
Top App Development Company
ASSOCHAM Member
Intended use decides it. Analyzing patient data to detect, diagnose, treat, or monitor a condition is SaMD.
510(k) needs a predicate and is faster. De Novo has none, takes longer, and sets a precedent.
Class II 510(k) needs analytical and clinical validation. De Novo needs more. A pre-sub confirms scope.
A Predetermined Change Control Plan covers post-clearance updates. We build PCCP-eligible architectures.
FDA review: 3–12 months for 510(k), 12–24 for De Novo. Build and prep add 6–18 months.
You do. Full IP transfer at close — code, data infrastructure, validation and submission docs.