EU AI Act · technical conformity evidence

Your AI audit, all from one tool.

Cartograph Suite generates the technical EU AI Act documentation (Annex IV/XI) for your AI model — fully automated, in a single run, as a finished dossier.

🔒 Runs locally on your side — your model weights never leave your machine.

Nine modules, one dossier

Each module delivers traceable technical evidence and is automatically mapped to the relevant articles and annexes of the EU AI Act.

M1

Cartography

Internal structure, layer/expert map, capacity concentration.

Interpretability · Art. 13/15
M2

Model profile

Architecture, exact parameter count, quantization, compute/memory estimate.

Annex IV / XI
M3

Leakage / memorization

Verbatim reproduction of known texts as a proxy for training-data/copyright leakage.

Annex XI · Art. 10
M4

Performance & calibration

Accuracy (logprob multiple-choice) plus calibration (ECE/Brier).

Art. 15 — accuracy
M5

Bias / fairness

Counterfactual tests over protected attributes, transparently scored.

Art. 10 — non-discrimination
M6

Robustness / red team

Jailbreak resistance, prompt injection (OWASP LLM01), perturbation stability.

Art. 15 · GPAI systemic risk
M7

Safety

Defensive refusal screening across harmful request categories.

Art. 5 / 15
M8

Transparency

AI self-disclosure under direct questioning and role-play pressure.

Art. 50 — labeling
M9

Doc generator

Aggregates everything into the finished compliance dossier with annex mapping.

The deliverable

As simple as that

No command line, no data uploads.

Choose a model

Select the folder with the weights or enter a model ID.

Start the audit

One click — all 9 modules run locally, trackable live.

Get the dossier

A finished, submittable EU AI Act dossier (Annex IV/XI) opens.

Sample dossier

Excerpt from a real run (model gemma-2-9b-it). Abridged — detailed tables, raw values and example outputs are reserved for the full dossier.

SAMPLE · SAMPLE
ModuleAreaFinding (verdict)
M1 CartographyStructure · Art. 13/15dense, FLAT — broadly used capacity
M2 ProfileAnnex IV/XI9.24 bn parameters · BF16 · [full architecture/compute table in the complete dossier]
M3 LeakageAnnex XI · Art. 10MEDIUM — only common knowledge, no long-form literature verbatim
M4 PerformanceArt. 15STRONG 92% · well calibrated (ECE 0.08)
M5 BiasArt. 10MEDIUM [attribute table in the complete dossier]
M6 RobustnessArt. 15 · GPAIWEAK — fiction jailbreak & prompt injection exploitable
M7 SafetyArt. 5/15ROBUST — 100% refusal across 7 categories
M8 TransparencyArt. 50MEDIUM — open when asked directly, deception under role-play pressure
M9 DossierAggregation12 sections · Annex IV/XI mapping · marked [automatic / by provider]

We produce the complete, submittable dossier for your own model — as a pilot or under your license.

Why it holds up

🔒

Local & offline

The tool runs on your machine. Weights and data never leave it — only the dossier comes out.

Verify before publish

Every finding is checked against the model's raw outputs before it's reported. Solid statements instead of blind values.

🧩

Across families

Validated across Llama, Mistral, Qwen, Gemma, Mixtral, Pythia — dense and Mixture-of-Experts alike.

📋

Clear responsibility

Every entry is marked: automatically collected vs. to be provided by the vendor. No grey areas.