Skip to content
Curious Context Labs

case study · Operations · Agent

Export paperwork done right the first time, without one person holding all the rules

How order and product data can produce commercial invoices, packing lists and carrier forms, with HS codes suggested and checked.

business
Logistics and distribution
build
Agent
timeline
6 weeks
read
3 min

01 the business

Specialist chemicals distributor

60 staff, shipping to 30 countries

Every international shipment needs its own documents with the same details typed again. A wrong weight, HS code or Incoterm can hold goods at the border. The rules live in one experienced coordinator's head.

02 before

The problem

Every international or regulated shipment needs its own set of documents, each with the same details typed again. One wrong weight, HS code or Incoterm can hold goods at the border or bring a fine. The rules sit in one experienced person's head, and nobody else is confident doing it.

  • Each document set takes an hour or more to prepare.
  • Mismatches between documents cause border delays.
  • Holidays and sick days create a backlog.

03 the build

What the AI does

  1. Step 1: Takes the order, product weights, dimensions and country of origin from NetSuite, SAP Business One or Cin7 when a shipment is booked.
  2. Step 2: Prepares the commercial invoice, packing list and carrier forms in the format the destination and carrier expect.
  3. Step 3: Suggests HS codes and flags items that may need extra documents, such as certificates of origin, safety data sheets or dangerous goods declarations, based on your product records.
  4. Step 4: Checks the paperwork against the order for mismatched quantities, values or addresses.
  5. Step 5: Saves the finished pack to SharePoint and attaches it to the shipment in ShipStation, ready for signature through DocuSign where needed.

paused: waiting for a person to approve

How it was put together

  • Took order, weight, dimension and origin data from the ERP when a shipment was booked.
  • Prepared commercial invoices, packing lists and carrier forms for the destination and carrier.
  • Suggested HS codes from product records and flagged items needing safety data sheets or dangerous goods declarations.
  • Checked all documents against the order and each other before saving the pack.

04 human in the loop

Where a person stays in charge

Your shipping or compliance lead checks and signs off every document set before it goes to the carrier or broker. New HS codes and any flagged items always need their review.

  • The compliance lead signs off every document set.
  • New HS codes and dangerous goods always need their review.
  • The system suggests; it never submits declarations to customs itself.

05 rollout

From kickoff to live in 6 weeks

  1. Weeks 1–3Document the coordinator's rules and clean product master data.
  2. Weeks 4–5Prepare packs in parallel and compare line by line.
  3. Week 6Switch to review-and-sign for standard destinations.

06 after

What changes

Paperwork is ready when the goods are, without retyping the same details. Fewer shipments get held up over avoidable errors, and the know-how no longer depends on one person.

worked example · your numbers will differ

Assume 300 international shipments a month at 60 minutes each to prepare by hand, falling to 15 minutes to review.

300 × 60 minutes = 300 hours a month today. 300 × 15 minutes = 75 hours. 225 hours a month saved.

What to measure

  • Time to prepare a document set.
  • Shipments held for paperwork errors.
  • Share of packs needing correction at review.
  • Number of people able to sign off shipments.

07 evidence

What the research says

  1. Gartner research put the average cost of poor data quality at $12.9 million a year per organisation.

    Gartner, Data quality: why it matters and how to achieve it (opens in a new tab) · 2020

  2. In a field experiment with 758 BCG consultants, those using GPT-4 on tasks within its abilities finished 12.2% more tasks, 25.1% faster and at over 40% higher quality, but did worse on a task outside its abilities.

    Harvard Business School / BCG, Navigating the Jagged Technological Frontier (Dell'Acqua et al.) (opens in a new tab) · 2023

08 lessons

Mistakes to avoid

  • Bad product data. HS codes and weights must be right in the master record.
  • Treating suggestions as decisions. Classification needs an accountable person.
  • Forgetting rule changes. Tariff and sanctions rules change often.

09 questions

Common questions

Can it classify goods automatically?
It suggests codes with the reason, but a qualified person confirms them. Classification errors carry real penalties.
Does it work with freight forwarders?
Yes. It prepares the pack in the format your forwarder or broker expects.

keep reading

All case studies ->

Tell us how your business runs. We'll tell you where AI fits.

A free 30-minute call. You talk, we ask questions, and you leave with a few concrete ideas — or an honest answer that AI isn't the right fix yet.

Working with teams around the world.