What is KlarvoEngine?

KlarvoEngine is Klarvo's regulatory intelligence engine. It analyses your AI system through a 3-pass classification pipeline and produces an article-cited compliance memo with obligations, deadlines, and recommended controls — in under 60 seconds.

What is the EU AI Act?

The EU AI Act (Regulation 2024/1689) is the world's first comprehensive AI regulation. It classifies AI systems by risk level and imposes obligations on providers and deployers of AI systems operating in the EU.

Any organisation that provides or deploys AI systems affecting people in the EU — regardless of where the organisation is based. This includes websites with AI chatbots, companies using AI analytics, and businesses integrating AI models into their products.

Prohibited AI practices have been enforceable since February 2025. GPAI model obligations since August 2025. Article 50 transparency and most remaining obligations apply from August 2, 2026.

What are the penalties?

Up to €35 million or 7% of global turnover for prohibited practices. Up to €15 million or 3% for transparency and other violations. For SMEs, the penalty is whichever amount is lower.

Adding Your First AI System

Powered by KlarvoEngine: When you add an AI system, KlarvoEngine automatically classifies it through a 3-pass pipeline. You describe the system in 4 fields — KlarvoEngine handles classification, obligations, and controls. See How KlarvoEngine Works.

This guide walks you through the complete process of adding an AI system to your Klarvo inventory — from preparation to the auto-generated compliance outputs.

Before You Begin

Gather the following information. You don't need everything upfront (the wizard creates tasks for missing items), but having these ready makes the process faster:

Information

Where to Find It

Required?

System name & internal ID

Your IT asset register or project name

Yes

Vendor/provider name

Contract or subscription records

If third-party

Primary owner

The person accountable for this system

Yes

Deployment regions

Where users or affected people are located

Yes

Affected groups

Customers, employees, candidates, students, etc.

Yes

Purpose description

What decisions or actions it influences

Yes

Vendor documentation

DPA, security docs, model cards

Recommended

Contract/terms link

URL or PDF of the agreement

If vendor-based

Starting the Wizard

Navigate to AI Systems in the left sidebar

Click the Add AI System button (top right)

Select your assessment mode on the first screen

Quick Capture Mode (2–4 minutes)

Best for initial inventory building, shadow AI triage, or low-priority systems. Captures essential identification, scope, and vendor info. Creates follow-up tasks for everything else. Classification is set to "Needs Review."

KlarvoEngine Classification (under 60 seconds)

Best for production systems, customer-facing AI, high-risk candidates, and audit preparation. KlarvoEngine's 3-pass pipeline handles prohibited practices screening, risk classification, transparency checks, data governance, human oversight, and FRIA trigger evaluation automatically. Produces a definitive classification memo with article citations in under 60 seconds.

Wizard Steps Overview

The wizard progresses through six logical phases:

Phase 1 — Foundation (Steps 0–3): Mode selection, basic info, vendor details, ownership assignment

Phase 2 — Scope & Context (Steps 4–5): Deployment regions, EU countries, affected groups, value chain role (deployer/provider/both)

Phase 3 — Classification (Steps 6–9): AI system definition test, use case description, prohibited practices screening (Article 5), high-risk screening (Annex III)

Phase 4 — Obligations (Steps 10–14): Transparency (Article 50), data & privacy, human oversight model, logging & records, incident response

Phase 5 — Special Contexts (Steps 15–17): Workplace-specific obligations, public authority context, training & AI literacy

Phase 6 — Finalization (Steps 18–20): FRIA trigger check, reviewer sign-off, completion summary

Each step includes contextual help text explaining why the question matters and what regulatory obligation it maps to.

After Completion

Once you finish the wizard, Klarvo auto-generates five key outputs:

AI System Record: A comprehensive database entry with all captured metadata, linked to your organization

Classification Memo (PDF): One-page auditable document showing inputs, classification result, confidence level, reviewer name, date, and rationale

Gap Checklist: A prioritized list of missing controls and evidence, grouped by obligation area (governance, transparency, oversight, logging, etc.)

Task Plan: Auto-created tasks with suggested owners and due dates — for example, "Upload vendor DPA," "Complete oversight SOP," "Assign AI literacy training"

Evidence Requests: Specific artifacts needed, tagged to the relevant control — for example, "Transparency notice screenshot required for TRN-01"

Tips for Success

✅ Be thorough: More detail in the wizard means fewer follow-up tasks and a higher-confidence classification

✅ Involve stakeholders: Loop in the actual system users — they know operational details you may not

✅ Attach evidence during the wizard: The wizard accepts file uploads at key steps — add vendor docs, screenshots, and policies as you go

✅ Use Quick Capture wisely: Great for building inventory fast, but schedule time to upgrade high-priority systems to KlarvoEngine Classification within 2 weeks

✅ Set realistic review dates: The default next review is 90 days after creation — adjust based on system criticality and change velocity