Overview

Use system_of_action when Klarefi should own the hosted intake flow.

Your backend creates an intake session, redirects the applicant to Klarefi, and then consumes runtime outcomes through webhooks or case reads. Klarefi handles the hosted intake experience, uploads, and gap resolution.

Typical Fit

• you want a hosted applicant-facing intake flow
• you want Klarefi to manage document collection and follow-up
• you want operators to review the resulting case in the same product

Basic Flow

1. Create an intake session.
2. Redirect the applicant to the signed hosted URL.
3. Receive events or poll case/session state.
4. Read the resulting runtime-backed case outcome.

Create A Session

POST /api/v1/sessions
Authorization: Bearer sk_live_...
Content-Type: application/json

{
  "case_type_id": "motor_claim",
  "idempotency_key": "session_2026_05_01_001",
  "external_case_id": "claim_12345",
  "ttl_hours": 168,
  "locale": "nl"
}

Example response:

{
  "session_id": "550e8400-e29b-41d4-a716-446655440000",
  "signed_url": "https://your-configured-app.example/s/550e8400...?token=...",
  "expires_at": "2026-01-22T10:30:00Z",
  "idempotent_replay": false
}

Redirect The Applicant

Send the applicant to the returned signed_url. From there, Klarefi owns the hosted intake flow, including structured sections, uploads, and gap-resolution prompts.

Consume Outcomes

After submission and processing, consume:

• webhook events for state changes
• case reads for the current runtime-backed result

The canonical result is the runtime-native case state and its shared projections, not a separate hosted-intake-only output model.