How a Logistics Office Assembles KYC Files Without Senior Hours

The workflow as it typically runs

Note: the description below reflects how this kind of work is commonly structured at logistics and freight-forwarding offices handling client and counterparty onboarding at volume. We are describing a genre-level pattern here, and we don't have visibility into your specific process.

A new client or counterparty submits documents: a Chamber of Commerce (KvK) extract, a UBO declaration, proof of identity, and for AEO-certified supply chains, sometimes additional customs paperwork. A compliance officer checks completeness against an internal checklist, screens names against sanctions and PEP lists, and chases missing items with the client. Once the file is complete, someone drafts a file summary for the senior who signs off, because an error here doesn't just damage the client relationship — it exposes the office's own obligations under the Wwft, the Dutch anti-money-laundering act.

In offices handling this at volume, the completeness check, the list screening, and the summary are where most of the time goes. That work is structured. It follows a defined logic, but it's currently done by hand, often at the end of the day when the senior finds time.

Where the method applies

The Orellis methodology begins with an operations map: every repeating task that takes more than 30 minutes per week, scored by frequency × pain × AI-reducibility.

In this file-assembly workflow, the tasks that score high on AI-reducibility share three characteristics:

• They follow a defined decision tree. "Is document X present in the file? If yes, proceed. If no, chase it." That's mechanical logic: easy to define, easy to check.
• They extract from structured or semi-structured inputs. KvK extracts and UBO declarations follow recognizable formats. Name, date of birth, role, and ownership percentage can be extracted reliably from a well-prompted AI system with a defined output schema.
• They produce a draft for the reviewer. The AI produces the first version of the completeness check, the list-match flag, and the file summary. The reviewer corrects and signs off before anything reaches a client, a register, or a supervisory authority.

The automation boundary we would draw

The most important decision in this engagement is where to stop. What to automate is the easy part.

AI handles:

• Document completeness check against a defined list
• Field extraction from the KvK extract and UBO declaration (name, date of birth, role, ownership percentage)
• Sanctions and PEP list comparison: a mechanical match that flags a possible hit, never clears one
• Draft file summary for the reviewer
• Draft client status update on missing items, for the reviewer to send

Reviewer retains:

• The final call on every sanctions or PEP match
• Edge-case interpretation
• The risk classification the Wwft requires of the office
• Sign-off on every file
• All communication with supervisory authorities

This boundary reflects both where AI is reliable and where the Wwft and the EU AI Act require documented human oversight for decisions with a regulated character.

What the build looks like

Orellis builds managed agents via Anthropic's Claude API, on infrastructure designed for EU data residency (no US data transfer required for the document content). The system runs against your intake data: submitted documents and form outputs. Every prompt and output is logged with a timestamp and operator ID.

The operator — your compliance officer or office manager — runs the system. Orellis builds it and hands it over. The runbook covers: how to run it, how to stop it, what to do when it returns an unexpected output, when to review the system prompt.

We scope the build during the engagement itself. We don't quote a fixed build fee until we've walked through your current intake process.

What this teardown is not claiming

• This is a description of how we would approach the problem, not a claim that we have delivered it for a client.
• Time savings and cost reductions depend on your current volume, document quality, and your reviewer's speed. We won't put a number on either until we've done the operations map.
• This adds a documented human-in-the-loop layer to your existing process. Compliance risk stays where it already sits: with the office, under the Wwft.

If you want numbers, the honest path is: run the audit, map the workflow, and measure. We can scope that conversation for free.