One guided EU AI Act workflow for every AI system.
From the first intake question to the last post-market review, every AI system follows the same path: registered, classified, assessed, documented and export-ready. Every step cites the Article it implements.
Full guidance from intake to signed conformity.
Confir auto-scopes the controls that apply, pre-fills what it already knows from your intake, and cites every step to the Article it implements — so the conformity package assembles itself as your team works.
Fast intake
Describe an AI system once — guided, plain-language, done in minutes.
Role & risk, classified
Provider or deployer, prohibited to minimal risk — determined against Article 6 and Annex III.
Scoped controls & obligations
Only the controls that apply to your role and risk class. Answers pre-filled from intake, gaps asked as plain questions, every high-risk duty surfaced.
Review & post-market monitoring
Human sign-off, then Article 72 monitoring that keeps its schedule — all written to an immutable Article 12 log.
Auto-classification and control scoping from a single intake.
The Act's text is encoded as rules, not inferred by a model. The same intake always produces the same classification, and every mapping cites the Article or Annex it came from.
Confir Intelligence
Rule engine — deterministic
Maps your description to every Article and Annex that applies.
Risk & role
High-risk · Provider
Art. 6 · Annex III
Auto-scoped controls
9 of 14 apply
Scoped from 22 obligations
Guided assessment
23 questions filled
Auto-filled from intake
High-risk docs
3 documents drafted
FRIA · Annex IV · DoC
Auto-classification — Use case, sector and supply-chain role mapped to Article 6 and Annex III in seconds.
Auto-scoping — Only the controls that apply to your system. No generic compliance checklists.
Pre-filled answers — System context, metadata and prior assessments carried forward automatically.
Answer once, use everywhere — Every datapoint flows through the assessment and into your documentation.
Initiation, Assessment, Review — one governed journey.
Every registered AI system moves through the same three phases. Sub-steps tick off as you work, findings appear in real time, and the next phase unlocks only when the previous one is complete — so nothing ships half-assessed.
- The same stepper governs every system
- Assessment progress tracked per compliance area
- Findings surface inline, with red badges
- Out-of-scope areas marked, not assessed
Resume Screening Model
Project Journey
0% overallOnly the controls that apply — pre-filled from your intake.
The assessment is not a blank questionnaire. Confir scopes the control catalogue to your exact role and risk class, pre-fills every answer the rules engine can derive, and converts every remaining gap into a rated finding with an owner and a due date.
Scoped by role × risk class
Out-of-scope obligations are filtered out, with the reason recorded.
Fully guided
Every control carries a plain-English brief. No need to read the Article first.
Pre-filled from intake
Derived answers arrive with their rationale attached.
Gaps become findings
Each gap is rated, routed to a remediation, and tracked to closure.
Assessment areas
Risk & Compliance
In progressAIRC · Art. 6 · 9
Question
Which EU AI Act risk class applies to this system?
Pre-filled from intake
High-risk — Annex III, point 1(a): biometric categorisation.
Transparency
CompleteAITR · Art. 13 · 50
Technical Documentation
In progressAITO · Art. 11 · Annex IV
Governance
Out of scopeAIGM · Art. 17
Answers the rules engine can derive are pre-filled from intake.
The AI System Datasheet: every fact in one record.
The Datasheet is the definitive record of an AI system — every field, stakeholder, classification and assessment, compiled from live platform data into one structured dossier. It's always current because it's never a copy, and it exports to a clean, print-ready PDF in one click.
- Compiled from live platform data — the record is never stale.
- Overview, stakeholders, classification, assessment, conformity — every section, in order.
- App-chrome-free, print-ready layout that exports to a clean PDF.
- One record to hand to an auditor, a customer, or your own board.
FRIA, Annex IV and DoC — generated from one workspace.
From a single system's assessment data, Confir assembles every mandatory high-risk document: the Annex IV technical file, the Fundamental Rights Impact Assessment, the EU Declaration of Conformity and the Article 50 Transparency Pack.
- Assembled from the control answers you've already given — no copy-paste, no duplicate data entry.
- Self-reconciling under Article 43(4): when a control answer changes, affected documents update.
- One click exports the complete set as Word or PDF.
CV & Recruitment Screening Tool
Annex IV Technical File
Art. 11 · 9 sections
Fundamental Rights Impact Assessment
Art. 27 · 7 sections
EU Declaration of Conformity
Art. 47 · Annex V verbatim
Art. 50 Transparency Pack
Art. 50 · 3 disclosures
Versioned under Art. 43(4) · Word + PDF in any of the 24 EU languages.
Every workflow capability, cited to its Article.
The capabilities that surround intake, assessment and conformity — each one cited to the Article or Annex it operationalises.
Systems Register & Lineage
Full lineage across every major and minor version.
Guided vs Expert Mode
Same data model, two ways in.
Auto-fulfilment
The rules engine pre-answers what your intake already implies.
Consolidation Phase
Formal review gate before conformity.
Received Declarations
Importer/distributor verification of upstream DoCs.
Multi-Language Export
Word + PDF in any supported EU language.
Datasheet & Reports
Print-ready, app-chrome-free views.
Self-Healing Reconciliation
Active projects auto-reconcile when controls evolve.
Assess your first AI system with Confir today.
Create your account, add your first AI system, and produce a full EU AI Act assessment with signed conformity documentation — in days, not months.