Check EPA origin documents before import filing
AI reads EPA origin documents for imports and checks the exporter, invoice number, goods, and quantity against the invoice. A customs specialist reviews issues.
- Runs on your own PCs and servers
- Use Azure OpenAI, Amazon Bedrock, Google Vertex AI, or an in-house model
- Its engine is Descarty's open-source Dagu (GPL-3.0)
The problem today
To claim a preferential rate under an economic partnership agreement (EPA), Japan requires the importer to submit or keep origin documents when the total customs value exceeds 200,000 yen. Each agreement sets its own required entries, and Japan Customs states that documents that don't meet them will not qualify except in limited cases. Before filing, the licensed customs specialist compares the origin documents with the invoice for missing entries and mismatches.
The workflow in Kitewell
- Receive the importer's email with the origin documents and invoice, and save the attachmentsEmail
- AI reads the agreement, type of proof, exporter, importer, invoice number, goods, HS code, and origin criterionAI judgment
- AI checks them against the agreement's required entries and the invoice, and shows possible gaps and mismatchesAI judgment
- The customs specialist reviews them and decides whether to claim the rate or ask the importerApproval
- Record the result and the reasons for the shipmentAutomatic
What AI does
Reads the needed fields from origin documents in any layout. Checks them against each agreement's required entries, set out as checking criteria, and against the invoice, and shows possible gaps and mismatches with where they appear.
Where people decide
Whether to claim the preferential rate, and whether an error can be treated as minor, are decided by the customs specialist in line with Japan Customs' rules. Unclear cases go to Customs for advice.
Prove Kitewell on your own systems
Try it free for 14 days on your own PC. To see it working first, ask for a 30-minute demo. There is also a one-month onboarding (¥500,000 before tax), where we build your first job together.
What to check when you try this work
- Whether the needed fields are read correctly from your most common agreements and proof types, such as third-party certificates and self-declarations
- How to set out each agreement's required entries and Customs' rules as checking criteria, and keep them in the project's knowledge
- How to ask importers to confirm, and how to keep the record
How the onboarding runs
- Week 1Requirements and success criteria
- Weeks 2 to 3Build
- Week 4Verify on production screens
- Final dayReport
One-month onboarding
- Length
- One month
- Cost
- ¥500,000 (tax excluded). Your staff build, with the developer beside them
- If you go on to a license
- The onboarding fee is credited to the first year's contract
- Without the onboarding
- At Team scale, start on Team monthly (¥14,800 per computer a month, tax excluded), and cancel in any month
You can try it first on your own PC, in the free 14-day trial.
Frequently asked questions
- Does AI decide whether the preferential rate applies?
- No. AI shows possible missing entries and mismatches between documents. Whether to claim the preferential rate, and whether an error can be treated as minor, are decided by the customs specialist in line with Japan Customs' rules.
- Do our documents and screens leave the company?
- Kitewell runs on your own PCs and servers, and Descarty never receives your business data. You choose the AI model: a cloud API, or an in-house or local model with an OpenAI-compatible API. With an in-house model, no data leaves the company. With a cloud model, what the AI reads is sent to the model's provider.
- What if the AI misreads something?
- A person checks what the AI produced before the work goes on. In a sheet, every value the AI reads shows the quote it came from, and rows below your confidence threshold go to a person. Reviewers can send work back with notes on what to fix, and who approved what is kept in the run history.
- What happens when a screen changes or a run stops?
- When a screen changes, the AI reads it again, relearns the step, and carries on. The relearned step is saved, and later runs repeat it. If a run still stops, ask the assistant right there, and it finds the cause and proposes a fix. A proposal is not saved until a person applies it.
- What happens to the RPA and systems we use today?
- Keep them. Kitewell operates screens the way a person does, so it works with systems that have no API. Move the robots that keep stopping and the work you gave up automating because it needs judgment, and leave fast, fixed processing in your current RPA.
- What does it cost?
- ¥14,800 per computer per month, or ¥148,000 per computer when paid yearly, before tax. A computer costs the same whether it builds or runs. You can try it free for 14 days. The one-month onboarding, where we build your first job together, is ¥500,000 before tax, credited to the first year if you go on.
Related use cases
Read and cross-check import documents before filing
AI reads the invoice, packing list, and B/L and cross-checks goods, quantities, weights, and values. A customs specialist resolves mismatches before filing.
Track vessels and tell shippers about ETA changes
Poll carriers' tracking APIs for your B/Ls and containers and detect ETA changes. AI drafts what the change affects and the shipper notice; staff check it.
Manage free time, demurrage, and detention
Calculate each container's remaining free time from discharge, gate-out, and empty-return events, and alert staff as deadlines near. AI drafts pickup requests.
See this work in a demo
A 30-minute demo shows this workflow running. The 14-day free trial lets you check it on your own systems.