SentinelEU AI Act
Log inStart free audit
About Sentinel

Building Trust
Through Evidence

Sentinel is an evidence-oriented compliance assessment platform — built to bridge the gap between regulatory obligations and observable technical reality.

Why Sentinel exists

The gap between regulation and code

Modern AI regulations introduce complex technical obligations that are difficult to interpret, verify and document consistently. Engineering teams work with source code and CI/CD pipelines. Legal and compliance teams work with regulatory texts and obligation frameworks.

The gap between these worlds leads to manual assessments, inconsistent interpretations and documentation that is difficult to reproduce or audit. Sentinel was built to bridge that gap through evidence-driven technical analysis.

Engineering team

Source codeCI/CD pipelinesInfrastructure

Compliance team

Regulatory textObligation frameworkLegal opinion

Sentinel bridges the gap

Evidence extractionArticle mappingSigned report

How Sentinel thinks

From artifacts to alignment

1

Software Artifacts

code · config · docs · CI/CD

2

Evidence Extraction

AST parsing · 33+ signal patterns

3

Finding Classification

severity · confidence · provenance

4

Article Requirements

22 EU AI Act articles

5

Alignment Scoring

weighted · deterministic · versioned

6

Compliance Report

signed · reproducible · Annex IV

Each step is traceable. Every finding links back to the artifact that produced it.

Our philosophy

Five principles that guide every decision

Evidence before opinion

Sentinel evaluates observable technical artifacts — source code, configuration, infrastructure, CI/CD pipelines — rather than questionnaires or subjective declarations.

Explainability

Every finding is traceable to the evidence that produced it. No black-box scoring. Each article score can be decomposed into its contributing signals.

Reproducibility

The same software analyzed under the same methodology version always produces the same result. Scores are deterministic, not probabilistic.

Transparency

Methodology, scoring weights and evidence models are documented and versioned. Changes to the rule engine are communicated before deployment.

Human oversight

Sentinel assists experts. It never replaces legal, engineering or regulatory judgment. Every report is designed to be reviewed and challenged by qualified professionals.

What Sentinel is not

×A legal opinion
×A conformity certificate
×A replacement for a Notified Body
×A black-box AI system

Built around evidence

What counts as evidence

Evidence in Sentinel is any observable technical artifact that can be linked to a regulatory requirement. Each evidence item carries its own provenance and confidence classification.

Full methodology
Source code structure and patterns
Configuration files and environment declarations
CI/CD pipeline definitions
Infrastructure-as-code and deployment manifests
Technical documentation and specifications
Package manifests and dependency graphs
Metadata and repository signals

Beyond the EU AI Act

A framework built to extend

Sentinel's architecture supports multiple regulatory frameworks through independent Rule Packs.

Active

EU AI Act

Regulation (EU) 2024/1689

Planned

ISO/IEC 42001

AI Management Systems

Planned

NIST AI RMF

AI Risk Management Framework

Compliance should not depend on assumptions.

Every assessment should be supported by observable evidence, transparent reasoning and reproducible methodology. Sentinel does not replace legal expertise — it provides the technical evidence that helps organisations make better compliance decisions.