Software Validation

Is Aligntra Validated?

A straightforward answer for regulated industries

The Short Answer

Yes, with a shared responsibility model. Aligntra is built with validation controls aligned to GAMP 5 and FDA CSA guidance. We validate the software; you validate its use at your site. We provide documentation to support your qualification activities.

What Kind of Software Is Aligntra?

Aligntra is a quality system support tool that helps your team review documents for regulatory compliance. It's designed to assist qualified personnel—not replace them.

            Important Distinction: Aligntra is NOT a medical device, NOT Software as a Medical Device (SaMD), and does NOT make autonomous compliance decisions. All AI findings require human review and approval.
        

GAMP 5 Category 4/5

FDA CSA Aligned

ISO 9001 Compatible

Quality System Software

What Validation Controls Are Built In?

✓

Automated Testing

227+ automated tests covering unit, integration, E2E, and security scenarios. Tests run on every code change before deployment.

⚙

AI Hallucination Detection

Every AI finding is checked against the source document. Semantic alignment scoring flags potential fabrications before they reach you.

🔍

Evidence Traceability

All findings link to exact source locations—page number, paragraph, and highlighted text. You can verify every AI output against the original.

📊

Confidence Scoring

Multi-factor confidence scores (HIGH/MEDIUM/LOW) help you prioritize human review. Low-confidence findings get flagged automatically.

🔒

Access Controls

Role-based access control, organization isolation, and enterprise-grade authentication via Clerk. Your data stays in your organization.

📋

Change Control

Git-based version control with 780+ tracked commits. Every change is documented, reviewed, and traceable to its author.

The Shared Responsibility Model

Like any SaaS platform in regulated industries, validation is a shared effort. Here's how responsibilities are divided:

◆ What We Do (Aligntra)

✓ Software development lifecycle controls
✓ Automated testing (227+ tests)
✓ AI output controls (hallucination detection)
✓ Evidence traceability to source documents
✓ Security controls and encryption
✓ Infrastructure and deployment validation
✓ Change management and release notes
✓ Validation documentation package

◆ What You Do (Your Organization)

✓ Determine intended use for your processes
✓ User Acceptance Testing (UAT)
✓ SOPs for how your team uses Aligntra
✓ User training on AI limitations
✓ Role assignments and access management
✓ Human review of all AI findings
✓ Record retention per your policies
✓ Final compliance determinations

Which Standards Does This Align With?

Our validation approach is designed to work across multiple regulated industries:

GAMP 5 Second Edition (2022) — Including guidance for AI/ML and cloud systems
FDA Computer Software Assurance (CSA) — Risk-based approach for production/QMS software
ISO 13485 — Medical device quality management systems
ISO 9001 — General quality management systems
AS9100D — Aerospace quality management
21 CFR Part 11 — Electronic records (see note below)

            21 CFR Part 11 Note: Basic audit logging is implemented. Organizations requiring full Part 11 compliance with tamper-proof audit trails should contact us about enterprise capabilities.
        

The Human-in-the-Loop Design

Aligntra is intentionally designed so that AI assists but never decides. This is critical for validation:

Every AI finding includes a source citation you can click to verify
Confidence scores highlight where human attention is most needed
You can accept, reject, or modify any finding before export
Final compliance determinations are always made by your qualified personnel

This design significantly reduces the validation burden compared to autonomous decision-making systems. The AI is a sophisticated search and analysis tool—your experts make the calls.

Validation Documentation We Provide

Validation Summary AVAILABLE

Overview of our validation approach, testing evidence, AI controls, and shared responsibility model.

21 CFR Part 11 Matrix AVAILABLE

Requirement-by-requirement breakdown showing vendor vs. customer responsibilities for each Part 11 control.

Release Notes AVAILABLE

Documented changes for each version, supporting your change control procedures.

Security Documentation ON REQUEST

Detailed security questionnaire responses and architecture documentation for your vendor qualification.

What We're Transparent About

Honest validation requires honest disclosure:

AI can miss things — False negatives are possible; AI supplements but doesn't replace expert review
AI can flag incorrectly — False positives happen; that's why human review is required
Confidence scores are estimates — They help prioritize review but aren't guarantees
Document quality matters — Poorly formatted documents may yield lower-quality analysis

These limitations are why we designed a human-in-the-loop system with full evidence traceability. Every finding can be verified against the source.

Need Validation Documentation?

We're happy to provide our validation summary, Part 11 compliance matrix, and answer your security questionnaire.

Request Documentation →