Regulatory pathway · all topics

Good Machine Learning Practice (GMLP)

Good Machine Learning Practice (GMLP) is a set of 10 guiding principles published jointly by FDA, Health Canada, and UK MHRA in October 2021. Principles cover multidisciplinary expertise, software engineering best practices, representative training data, independence of training and test sets, reference standards, model design, performance focused on human-AI team, performance testing in clinically relevant conditions, clear information to users, and post-deployment monitoring. Applies to ML-enabled medical devices and SaMD.

Source

FDA, Health Canada, MHRA, 'Good Machine Learning Practice for Medical Device Development: Guiding Principles' (Oct 2021)

https://www.fda.gov/medical-devices/software-medical-device-samd/good-machine-learning-practice-medical-device-development-guiding-principles

How MedTek OS handles this

Compass drafts the good machine learning practice (gmlp) workflow with confidence-gated content.

The platform refuses to generate content below the threshold for this artifact class. Every claim carries a citation chain back to RIG, audit-logged with HMAC integrity.

Try the sandbox Ask the agent about this

Related pathways

510(k) Premarket Notification

A 510(k) clearance under 21 CFR Part 807 Subpart E demonstrates substantial equivalence to a legally marketed predicate device. Required for...

De Novo Classification Request

De Novo creates a new classification for a novel low-to-moderate risk device with no valid predicate. Submission under 21 CFR Part 860 Subpa...

Premarket Approval (PMA)

PMA is the most stringent FDA pathway, required for Class III devices that support or sustain human life or pose significant risk. Filed und...

Investigational Device Exemption (IDE)

An IDE under 21 CFR Part 812 allows an unapproved device to be used in a clinical study to gather safety and effectiveness data. Required wh...