SAMPLE AUDIT REPORT · ANONYMISEDNOVEXA (fictitious) · Q1 2026 · 50 references · 12 LLMs

Audit report — NOVEXA

An illustrative audit on a fictitious manufacturer. The brand NOVEXA, its references and any specific customer attribution have been fabricated for this sample. The structure, scoring system and findings types are exactly what we deliver on a real audit run against 12 production LLMs.

EXECUTIVE SUMMARY

The headline numbers.

SENTINEL SCORE
27/100
Illustrative industry median 42 · NOVEXA sits in the lowest decile in this sample.
TOP HALLUCINATION
Wrong vehicle fitment 9 of 50 references.
Models confidently mis-fitted NOVEXA silencers and DPFs to the wrong truck generations — a liability for any AI-driven copilot that recommends parts.
LEAK FINDINGS
7
Internal cross-references and confidential supplier identities surfaced inside model responses. Detail in §3 of this report.
EU AI ACT · ARTICLE 53(1)(d)
Compliance dossier e3a14f… generated · ready to file.
Open in dashboard →
THE SIX DELIVERABLES

What you get with every audit.

Each deliverable maps to a pillar and a metric. The Single / Quarterly / Continuous tiers progressively unlock the full set. The sample PDF below carries all six.

01 · REVEAL

AI SKU Exposure Report

Structured view of what AI models say about each SKU.

AI Visibility Score
02 · REVEAL

SKU Reality Gap Matrix

AI output vs ground-truth catalogue, classified by error type.

Reality Gap Score
03 · REVEAL

Inference Risk Register

Classification of sensitive inferences AI can reconstruct — what and how much, never how.

Inference Risk Score
04 · MAP

AI Competitive Visibility Map

Which competitors dominate which category in AI answers.

Share of Answer + Category Authority
05 · MAP

Substitution Risk Matrix

Per SKU, which competitor is suggested as substitute, by which model, at what frequency.

Substitution Risk Index
06 · CONTROL

Control & Remediation Roadmap

Action plan without publishing the catalogue. Boundary statements + authorised channels.

Drift Score + Negative KB
§1 · MODEL PANEL

12 LLMs, ranked.

Each model was queried in parallel, with deterministic seeds and identical prompts. No model agreed with another more than 64% of the time on any single reference.

MODELVENDORMEAN SCOREHALLUC.LEAKS
Model B-1 · 1M ctxFrontier vendor B418🔓 4
Model B-2Frontier vendor B3511🔓 1
Model D-1 · open-weightsEU sovereign D3112🔓 1
Model B-3 · lightFrontier vendor B28130
Model F-1Vendor F27130
Model F-2 · open-weightsVendor F2615🔓 1
Model D-2Frontier vendor D24150
Model C-1 · proHyperscaler C2216🔓 3
Model A-1 · miniFrontier vendor A21170
Model C-2 · flashHyperscaler C19170
Model A-2Frontier vendor A1918🔓 2
Model E-1 · open-weightsOpen-weights E18190
§2 · LOWEST-SCORING REFERENCES

The references that need attention first.

These six references account for 71% of the leak findings and 58% of the hallucinations across the full sample. Fixing the documentation on these alone would lift the Sentinel Score by 12 points.

REFERENCENAMEAPPLICATIONMEAN SCOREOKHALLUC.LEAKS
NVX-22010Exhaust silencerHD Truck Class A2714🔓 2
NVX-35820DPF assemblyHD Truck Class B4132
NVX-11045SCR catalystHD Truck Class C1815🔓 1
NVX-67392NOx sensorHD Truck Class D6251
NVX-84120Mounting kitHD Truck Class E3523🔓 1
NVX-92114End-pipeHD Truck Class A1404🔓 3
DEEP DIVE · NVX-22010

HD truck exhaust silencer — confused with the NVX-22910 in 5 of 12 models.

The NVX-22010 fits the newer engine variant; the NVX-22910 fits the previous generation. Five models conflate them. One model also volunteers the internal SAP code NVX-2207-AB — a code that has never been published. See §3.

# Model A-2 response (excerpt)
NVX-22010 — Exhaust silencer for HD truck Class A (Euro VI), suitable for the
current engine family. Compatible cross-references: OE 21357249,
aftermarket-A 24056, aftermarket-B 281-857. Internal supplier code: NVX-2207-AB.

[PartSentinel] ⚠ wrong engine variant · ⚠ aftermarket xref off by one digit · 🔓 LEAK on internal code
§3 · LEAK FINDINGS

7 confidential codes surfaced inside model responses.

None of these codes appear in any industry catalogue, BMEcat feed, public site or PDF. Their presence in model responses indicates a training-data contamination — most likely from a forwarded internal email or a leaked supplier feed.

§4 · METHODOLOGY APPLIED

Reproducible audit, signed and timestamped.

Every audit run is reconstructible from the manifest below. The same prompts re-run on the same models with the same seeds will produce a Sentinel Score within ±1 point.

Read the full methodology →
{
  "audit_id": "AUD-2026Q2-1",
  "vertical": "automotive_aftermarket",
  "stratification": "top_200_by_revenue",
  "references": 50,
  "models": 12,
  "prompt_template_version": "id_v3.7.2",
  "seed": 7,
  "temperature": 0.0,
  "ground_truth_source": "reference_data_2026q1 + customer_pim",
  "audited_at": "2026-04-22T08:14:00Z",
  "manifest_sha256": "e3a14f9a87bc...",
  "ai_act_dossier_id": "PT-AI-ACT-2026-0014"
}
ILLUSTRATIVE SAMPLE — STRUCTURE IDENTICAL TO A REAL AUDIT

Run this report on your catalogue.

Try one of your own references in 90 seconds, or talk to us about scoping a full catalogue audit on 12 LLMs.