AI Readiness Agent

The AI Readiness Agent is Ariftly's flagship agent. It automates EU AI Act compliance analysis, answers enterprise procurement questionnaires, and audits your AI knowledge base — cutting weeks of manual work down to hours.

What it does

Capability	Description
Procurement Questionnaire Response	Ingests 40-page enterprise AI/security questionnaires and drafts complete answers grounded in your knowledge base
EU AI Act Gap Analysis	Maps your AI systems against EU AI Act requirements, identifies gaps, and prioritizes remediation
Knowledge Base Audit	Reviews your AI policies, data governance docs, and technical documentation for completeness
Risk Classification	Classifies AI systems by risk tier (prohibited / high / limited / minimal) per the EU AI Act
Compliance Report Generation	Produces board-ready compliance reports as downloadable artifacts

Getting started

1. Connect your knowledge base

The agent grounds all its reasoning in your documentation. Upload or connect:

• AI governance policies
• Data processing agreements
• Model cards and training data documentation
• Existing compliance certifications (SOC 2, ISO 27001, etc.)

Go to Dashboard → Knowledge → AI Readiness to add documents.

2. Deploy the agent

# Via API
curl -X POST https://api.ariftly.io/v1/agents/ai-readiness/deploy \
  -H "Authorization: Bearer $ARIFTLY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"plan": "starter"}'

Or click Deploy on the AI Readiness Agent card in the dashboard.

3. Trigger your first task

Questionnaire response:

curl -X POST https://api.ariftly.io/v1/tasks \
  -H "Authorization: Bearer $ARIFTLY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "agent": "ai-readiness",
    "task_type": "questionnaire.respond",
    "input": {
      "document_url": "https://...",
      "format": "pdf"
    }
  }'

Gap analysis:

curl -X POST https://api.ariftly.io/v1/tasks \
  -H "Authorization: Bearer $ARIFTLY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "agent": "ai-readiness",
    "task_type": "compliance.gap_analysis",
    "input": {
      "framework": "eu_ai_act",
      "scope": "all_ai_systems"
    }
  }'

Task types

Task Type	Input	Output Artifact
questionnaire.respond	Document URL or text	ai_readiness.questionnaire_response
compliance.gap_analysis	Framework, scope	ai_readiness.gap_analysis
knowledge.audit	Collection ID	ai_readiness.audit_report
report.generate	Gap analysis artifact ID	ai_readiness.compliance_report

Approval gates

The agent requests approval before:

• Sending a questionnaire response to any external recipient
• Publishing a compliance report to a shared channel

All other work (analysis, drafting, auditing) runs autonomously and surfaces findings in the dashboard.

Integrations

Works best with:

• Slack — receive findings and approve actions without leaving Slack
• Gmail — forward questionnaire emails directly to the agent
• GitHub — reference your repository's AI-related code and documentation

Extending with Skills

Add custom behavior without code:

id: ai_readiness.escalate_high_risk
agent: ai-readiness
triggers:
  - on: artifact.emitted
    when:
      artifact.type: ai_readiness.gap_analysis
      artifact.data.risk_tier: high
actions:
  - type: notify
    target: { role: "cto" }
    channel: slack
    message: "High-risk AI Act gap identified — review required"
  - type: create_approval
    title: "Escalate to legal team?"

→ Learn about Skills