Turn fragmented AI documentation into structured, reviewable evidence in 72 hours. Defined-scope engagements for Annex IV preparation and internal review.
No commitment required. Scope is confirmed first.
EU AI Act enforcement begins August 2, 2026, with evolving timelines under current regulatory proposals. Most organisations remain unprepared for structured evidence requirements.
AnnexLayer prepares structured evidence outputs before formal audit, legal validation, or governance platform onboarding. It is not a consulting service. It is not legal advice. It is not a certification provider. It is the layer that turns fragmented documentation into usable evidence before external review begins.
Before committing to a full governance platform, an extended advisory engagement, or an internal build programme, most teams need one thing first: a clear, bounded evidence output that shows where the gaps actually are.
That is what AnnexLayer delivers. One AI system. Defined inputs. Structured output. 72 hours.
Before engaging consultants or advisory firms
Before internal audit review or governance board sessions
When documentation exists but is not structured or reviewable
When timelines are compressed and evidence is needed quickly
When you need a fast evidence baseline for internal decision-making
Professional human-readable PDF with executive summary, field-by-field findings, gap identification, and remediation context.
Machine-readable structured evidence object with all extracted fields, suitable for governance platforms, internal systems, and audit ingestion.
SHA-256 cryptographic hash of the complete evidence payload. Timestamped records support EU AI Act Article 26 logging and retention workflows.
Complete corrections log with reviewer ID, UTC timestamp, and decision rationale for every reviewed field.
Every engagement covers one AI system with defined documentation inputs, structured evidence output, and human-reviewed release. Scope is confirmed before work begins. Larger environments move into extended scope or portfolio pricing.
Defined-scope 72-hour evidence workflow for one AI system. Identifies Annex IV documentation gaps and produces structured evidence outputs. The bounded starting point before deeper work.
Follow-on systems move into Full Evidence Pack pricing.
Complete Annex IV structured evidence documentation with audit trail and gap analysis. Deeper extraction, broader field coverage, and remediation priorities.
Gap analysis against EU AI Act requirements with prioritised remediation insights. The “are we ready before external review?” engagement.
Ongoing monitoring, delta tracking, and continuous evidence updates. Monthly re-scan of your AI vendor ecosystem with change reporting.
Use the evidence outputs to brief your governance team, board, or risk committee on current AI system documentation status.
Provide the structured evidence pack to legal counsel, external auditors, or advisory firms as a prepared input for their review.
Use the gap analysis and remediation context to prepare for internal audit, readiness review, or regulatory engagement.
Determine whether to remediate internally, expand into additional systems, or engage further support based on clear evidence.
Every data point is independently extracted by two separate engines. Disagreements are flagged for review — never silently resolved. Full transparency on evidence agreement and divergence.
No output is released without human review. Every correction is recorded with reviewer ID, UTC timestamp, and decision rationale. Complete corrections log included with every deliverable.
All evidence outputs include timestamped audit logs, reviewer actions, and structured traceability. Records are retained to support EU AI Act Article 26 deployer logging obligations and audit workflows.
Every engagement produces a complete audit trail including extraction records, human review actions, and final evidence states. Outputs are structured to support minimum retention expectations under EU AI Act deployer obligations.
No client data is used for AI model training. All processing is configured with explicit training data opt-outs and controlled data handling.
One AI system, defined documentation inputs, structured evidence output (PDF report + JSON pack), and human-reviewed release within a 72-hour delivery window.
One distinct AI application, model, or workflow with a defined intended purpose. If a vendor bundles multiple AI capabilities under one product name, scope is assessed and confirmed before work begins.
Environments with multiple systems or extensive documentation move into extended scope or portfolio pricing, discussed and agreed before engagement starts.
The delivery window begins when documentation inputs are received and confirmed as complete.
No. Outputs are structured evidence drafts designed to support internal review and external advisory processes. They do not constitute legal advice, certification, or formal regulatory determination.
Most teams move into Full Evidence Packs for additional systems or Evidence Maintenance for ongoing monitoring and delta tracking.
No. All processing is configured with explicit training data opt-outs. Client documentation is never used for model training and never shared with third parties.
EU AI Act enforcement begins August 2, 2026, with evolving timelines under current regulatory proposals. A realistic compliance runway is 32–56 weeks for most mid-market organisations.
Yes. Every engagement includes timestamped audit logs, reviewer decisions, and structured evidence outputs designed to support EU AI Act Article 26 deployer obligations.
Important disclaimer: AnnexLayer provides structured evidence drafts designed to support internal review and external advisory processes. Our deliverables do not constitute legal advice, certified assessments, formal conformity determinations, or regulatory certification. All outputs are clearly labelled as structured evidence requiring client validation and appropriate professional review.
