Version: 1.1.0

This guide translates traditional GxP compliance milestones into the ROSIE framework components. It is designed for QA, Regulatory Affairs, and auditors to understand how digital evidence replaces paper artifacts.


Understanding ROSIE’s Boundary

ROSIE is a specification, not a complete system. It defines:

  • How to structure compliance artifacts in code repositories
  • How to compute integrity hashes
  • The API contract for approval systems

Your organization’s System of Record (SoR) handles:

  • User authentication and authorization
  • Approval workflows and electronic signatures
  • Audit trail storage and retention

Any QMS, PLM, or custom system that implements the ROSIE API (RFC-004) can serve as the SoR.


1. Traditional vs. ROSIE Mapping

Traditional GxP DeliverableROSIE Digital EquivalentWhy This is Better for Compliance
User Requirements (URS)specs/urs/*.mdLiving specs: requirements are version-controlled alongside code
Functional/Design Specs (FRS/DS)specs/ds/*.md and @gxp-idDirect links: the design is part of the code
Traceability Matrix (RTM)ROSIE graph (auto-generated)Real-time accuracy: derived from code, no manual drift
Test Protocols (IQ/OQ/PQ)tests/ with @gxp-typeAutomated execution: tests run on every change
Validation Summary Reportgxp-execution.jsonImmutable evidence: logs, hashes, and screenshots
Handwritten signaturesSoR e-signaturesPart 11 compliant: signatures are linked to a specific Git hash

2. Navigating the Audit Trail

In a traditional audit, you look at a binder. In a ROSIE-enabled audit, you look at the Dual-Ledger.

1. Prove the Intent (the repo)

Auditor question: “Where did you say the system would do X?”

Answer: The Markdown file in /specs. Git history shows who wrote it and when.

2. Prove the Design (the code)

Auditor question: “How did you implement that requirement?”

Answer: The ROSIE Engine follows the @gxp-id tag to the exact function in code.

3. Prove the Verification (the evidence)

Auditor question: “How do I know this actually works?”

Answer: The gxp-execution.json shows test results, evidence attachments, and the RRT.


3. Hard-Gate vs. Traditional “Go-Live”

Traditional “Go-Live”ROSIE Hard-Gate
Human check: a person verifies signed documentsDigital check: deployment is blocked if any signature or hash is missing
Point-in-time: validation is a snapshot from last monthContinuous: validation is a live status of the current build
High friction: releases wait on paperwork routingLow friction: releases proceed because evidence is generated during development

4. Addressing GAMP 5 and CSA Principles

  • Critical thinking (CSA): The AI agent flags high-risk changes (e.g., security, data integrity), so QA focuses on the most important areas
  • Fitness for purpose: Self-validation proves the tool works correctly on every run, not just during a one-time validation

5. Summary for Auditors

ROSIE supports 21 CFR Part 11 compliance by ensuring:

  • Integrity: The manifest hash prevents undetected changes to validated code
  • Traceability: Every requirement is digitally linked to code and tests
  • Accountability: Every approval is a timestamped, cryptographically secure e-signature

Key References

DocumentPurpose
RFC-001: Data StandardHow requirements and traces are structured
RFC-002: Engine SpecHow the Hard-Gate and sync work
RFC-003: Evidence StandardHow evidence is captured and packaged
RFC-004: API InterfaceHow ROSIE connects to your SoR
RFC-005: TQ BaselineHow the tool validates itself