UK Customs Declaration Automation for Freight Forwarders

Introduction

How Ripple automates UK customs declaration preparation for freight forwarders, from document intake and commodity code validation to CDS-ready entries in ASM Sequoia and CargoWise.

Date
November 17, 2026
UPDATED
November 17, 2026
Author
Jonny Smith
Type
Guide

Ripple automates the full customs declaration preparation workflow for UK freight forwarders: from document intake through commodity code validation, CDS data preparation and exception routing, to a complete audit trail behind every decision. It handles imports and exports. Experienced customs staff retain control of submission. Ripple handles everything before it.

At Pentagon Freight Services, Ripple removes up to 1,400 hours a week of manual UK customs preparation across around 140 declarations a day, delivering a 60% reduction in clearance costs, 40% lower software costs and 99%+ first-pass classification accuracy on live shipments.

This page sets out what Ripple does at each stage of the workflow, how it handles imports and exports differently, how it integrates with ASM Sequoia and CargoWise, and what the results look like in live UK production.

What Ripple Does Across the Customs Preparation Workflow

The submission to HMRC’s Customs Declaration Service (CDS) is rarely the bottleneck. The bottleneck is everything that happens before an operator opens the declaration software: document triage, data extraction, rekeying, commodity code logic and pre-submission checking. Each stage is a source of errors, delays and wasted senior broker time.

The scale makes it worse. A single industrial entry can run to hundreds or thousands of line items, each needing a code, value, weight and origin. Without automation, more volume means more staff.

Ripple addresses each stage as a distinct operational problem, not as a single extraction task.

Document intake and triage

Ripple monitors the mailboxes where customers send customs paperwork, around the clock, and opens a job for every request. It identifies what has arrived, classifies documents by type, associates them with the correct shipment, and flags incomplete packs at the point of arrival rather than mid-preparation.

  • Job details from the email. The subject line carries the job reference, transport mode, branch and direction (import or export). Defaults can be set per customer.
  • Instructions in the body take priority. Gross weight, packages, valuation method, nature of transaction, commodity codes or party details in the email body override what is in the documents.
  • CargoWise packs arrive ready. Packs sent automatically from CargoWise come with the correct subject header already set.
  • Structured feeds too. Ripple can also take client-specific templates and feeds, such as a shipper’s manifest triggered when a warehouse closes a container.

For forwarders handling large numbers of customers at once, this removes the inbox management burden from customs staff, and missing documents surface before an entry has been started, not after.

Data extraction without templates

Ripple extracts structured, line-level data from every document in the pack: PDF (native and scanned), XLSX, XLSM, XLS, DOCX and RTF. It does not need a separate template for each customer’s document layout.

  • Commercial invoice: item number, product code, description, quantity, line value, invoice number and date, currency and parties.
  • Packing list: item-level gross and net weight, package count and country of origin, matched back to the invoice by item number.
  • Clearance instruction, booking sheet or shipper’s letter of instruction (SLI): job-level details and instructions.
  • Customer summary document: where supplied, the customer’s approved customs description and commodity code per item.
  • Transport documents: master and house bill of lading or air waybill numbers and dates.

It handles the real-world mess: line items split across merged cells, multi-page and multi-document packs, duplicate invoices, and packing-list pages mixed in with invoices.

This matters operationally. Template-dependent tools need a build cycle before each new customer can be onboarded, and ongoing maintenance whenever a customer changes their invoice format. Ripple generalises across formats, so new customers are onboarded without a template build, and logic changes are deployed centrally rather than template by template.

For detail on the five stages where manual data entry occurs and what each costs, see The 5 Stages of Manual Data Entry Before a CDS Declaration.

Applying each customer’s own rules

This is where Ripple differs from a lookup tool. Every customer builds declarations in their own way, and HMRC has its own requirements. Ripple encodes both as rules, so the same job always produces the same answer, whichever operator is working the queue. Rules running in UK production include:

  • Document priority. When a pack holds several documents, use them in a set order, for example clearance instruction first, then SLI, then invoice.
  • Fallbacks. If the primary document is missing weight, packages, value, parties, description or code, look in the next document down.
  • Which source wins for which field. The invoice is authoritative for quantity and value, the packing list for weights, the customer’s summary sheet for description and commodity code.
  • Item-number matching. The item number ties invoice, packing list and summary sheet together.
  • Duplicates and non-goods. Invoices with the same number are extracted once. Packing materials and other non-goods lines are removed, and the user is told what was removed and how to add it back.
  • Grouping. Consolidate lines by product code, commodity code, procedure code and country of origin, summing values and weights within each group, or keep every line separate when the job says “do not group”.
  • Weight apportionment. If item-level weights are missing, allocate the total gross weight in proportion to each group’s share of invoice value.
  • Party matching. Match importer and exporter names to the customer’s address book and customs records to fill EORI numbers and addresses, and flag when there is no match.

Commodity code validation and classification

Ripple assigns or checks the commodity code for every line item, applying UK Global Tariff (UKGT) classification logic and each client’s own approved rules. It follows the same approach a trained classifier would: read what the product actually is, place it in the right heading using the tariff structure and the General Interpretative Rules, then narrow down to the national subdivision.

  • Customer catalogue first. If a product has been classified and confirmed before, the confirmed code is reused instantly.
  • AI classification for new items. New or changed products get a suggested code with a confidence score and the reasoning behind it.
  • Validation. Every code is checked for the correct length for the direction (10 digits on a UK import, 8 on an export), for validity, and against the product description.
  • Origin and preference checks. Procedure codes, origin evidence and preference claims are checked before the entry reaches CDS, not after a rejection forces a rework cycle.
  • People decide the hard ones. Borderline codes, missing origin evidence and unusual product combinations are routed to a named reviewer with a suggested answer and the specific issue identified. Routine items proceed automatically.
Why this matters under CDS Release 5.1.0: HMRC’s release on 28 March 2026 tightened validation of DE 5/15 (country of origin) and DE 5/16 (country of preferential origin). Pre-lodged declarations that would previously have passed are now rejected if the two fields and the preference code do not align. Ripple validates these fields at intake, before submission.

Classification logic is managed centrally, not held in individual operators’ heads or per-customer spreadsheets. Confirmed decisions feed back into the customer catalogue, so the next match is instant.

For a full account of how CDS validation changes are affecting pre-lodged declarations, see UK Customs Classification and CDS Compliance: How to Reduce Clearance Rework.

Imports and Exports: Different Jobs, Handled Differently

Imports and exports need different data, different codes and different CDS declarations. Every job Ripple processes is tagged with its direction before anything else happens. If the direction is missing, Ripple asks for it rather than guessing.

On UK imports, Ripple:

  • Expects a 10-digit commodity code (digits 9 and 10 can change the duty and measures applied)
  • Requires a valuation method (codes 1 to 6) and country of origin at item level
  • Checks the importer EORI, customs value, nature of transaction, procedure code (CPC), weights and packages

On UK exports, Ripple:

  • Expects an 8-digit commodity code
  • Checks the exporter EORI and the consignee name and address
  • Checks nature of transaction, CPC, weights and packages, with country of origin optional

The common failure points differ too. Imports typically fail on a missing valuation method or origin, the wrong code length, or a party that cannot be matched to an EORI. Exports typically fail on the wrong code length or a missing exporter EORI or consignee address. Ripple checks for each before the entry is built.

Direct Integration with ASM Sequoia and CargoWise

Ripple writes structured, CDS-compliant data directly into the declaration system your team already uses. For forwarders on ASM Sequoia, entries are created in Sequoia-ready format from source documents. For CargoWise operations, Ripple integrates two ways: it pushes classification and declaration data in and pulls job data out, working alongside the system rather than replacing it.

This removes the rekeying step entirely. Extracted and validated data arrives in the target system against the correct job, with header information (invoice value, invoice number and date, transport document references), complete invoice rows, and any approved consolidation applied. Every line is checked back against the source documents before it goes anywhere. The operator does not transfer values by hand.

What integration acceptance looks like in practice:

  • A representative document pack arrives in the receiving system against the correct job
  • Header information, invoice rows and approved consolidation are present
  • Flagged exceptions are visible and actionable within the operator’s existing workflow
  • The operator can inspect, resolve exceptions and submit without leaving their declaration system

Pre-Submission Validation and Exception Handling

Ripple validates data against CDS requirements before the entry reaches submission. Standard entries flow to the operator ready to check and submit. Genuine exceptions are routed with the specific issue identified, so the operator’s attention goes where it is actually needed.

Every job gets a clear outcome: successful, successful with information (for example, which non-goods lines were removed), a warning (data missing) or an error (data invalid), each stating exactly what to send back.

Where data is missing, Ripple chases it

Ripple replies to the sender asking for the specific field, in a format they can answer in one line, and carries on when they reply. A typical exchange from live UK operations (references anonymised):

Ripple: Missing Valuation Method (imports only). Please reply with: Valuation Method: X (valid codes 1-6)
Sender: Valuation Method: 1
Ripple: Missing commodity code for 2 items.
Sender: Commodity Code: 8479899790 - Item: SELF-ALIGN MULESHOE GUIDE. Commodity Code: 7003199000 - Item: Glass
Ripple: The outcome of your submission is successful. Info: 3 packing items were removed as non-goods. To add any back, reply with Commodity Code: XXXXXXXX - Item: Description

Exceptions handled this way include missing gross weight or package count, missing valuation method or nature of transaction code, a missing, invalid or wrong-length commodity code, missing CPC, value or country of origin, missing importer or exporter EORI, name or address, no match in the customer’s address book, a missing job reference, mode or direction, and no goods lines found.

The pre-submission check changes in character. The operator is reviewing a validated, AI-prepared entry rather than compensating for accumulated manual errors.

Full Audit Trail for Every Decision

For every entry Ripple prepares, the audit trail records:

  • The source documents used (invoice, packing list, bill of lading)
  • The commodity code recommended, with the classification logic applied
  • The interpretative rule (GIR 1 to GIR 6) referenced
  • Any alternative codes considered and why they were excluded
  • Whether a human reviewer accepted or overrode the recommendation, and who
  • The override reason where an AI recommendation was not accepted
  • The timestamp and reviewer identity for every decision

This evidence is generated automatically as part of the workflow and is visible in Ripple’s bespoke user interface, alongside every job, the extracted data per line and the items awaiting review. Under HMRC’s archiving requirements, classification records must be kept for four years and be readily accessible for inspection. Ripple produces a complete, searchable audit trail for every shipment without adding a documentation burden to the clearance team.

For the full evidence standard HMRC expects behind every classification decision, see How to Prove Why a Commodity Code Was Selected.

Operational Results: Pentagon Freight Services

Pentagon Freight Services is a global freight forwarder with a large UK operation, handling thousands of complex clearances a month. Ripple runs the end-to-end customs preparation workflow in live UK production: from commercial invoice receipt through to CDS submission-ready declarations in ASM Sequoia and CargoWise.

These are not pilot results. Pentagon runs Ripple across real CDS submission volumes, real multi-document complexity and real compliance requirements every day.

UK results at Pentagon Freight Services:

  • Around 140 declarations prepared a day, including large multi-line, multi-document entries
  • Up to 1,400 hours a week of manual customs preparation removed, equivalent to up to 35 FTEs of capacity
  • 60% lower clearance costs
  • 40% lower software costs
  • 99%+ first-pass classification accuracy on live shipments, with classification cycles cut from weeks to seconds
  • 1,000+ line entries checked as correct in seconds
“Large complex entries were coming out perfect.” — UK Customs Operator, Pentagon Freight Services

The 1,400 hours reflects the preparation time that previously fell on Pentagon’s customs team before each declaration reached submission. At up to two hours of preparation per declaration across five working days, that work is now handled by Ripple. Pentagon’s customs team checks and submits.

Beyond the UK. Ripple is also in production in the UAE, where Pentagon selected Ripple for customs and warehouse operations in Dubai, and in the US. The approach is the same; the codes, data and systems are configured per market. See our UAE customs automation guide. The results above are UK results.

Who Ripple Is Built For

Ripple is designed for UK freight forwarders and customs brokers who are:

  • Processing high declaration volumes where manual preparation is the main capacity constraint
  • Handling complex, multi-line industrial entries, such as oil and gas specialist equipment, where a single entry can run to hundreds or thousands of lines
  • Running ASM Sequoia or CargoWise and need automation that writes directly into their existing system, not a separate tool
  • Managing diverse customer document formats that make template-based extraction tools impractical to maintain at scale
  • Under pressure from CDS validation tightening, particularly around origin, preference and pre-lodged declarations
  • Needing an auditable record of every classification and declaration decision for HMRC compliance and internal governance

Ripple goes live in weeks. Each deployment is built around the client’s own customs rules, customer base and declaration workflow. It does not ask teams to change how they work.

What Ripple is not

Ripple is not a self-serve SaaS tool that sits alongside your existing workflow, and it is not a licensed customs broker. It is an AI operations layer that owns the preparation workflow end-to-end, designed, built and continuously improved by Ripple’s team for each client’s operation. Your licensed team stays accountable for submission. That model is why the results at Pentagon are production outcomes, not proof-of-concept numbers.

Frequently Asked Questions

Does Ripple prepare data for CDS declarations?

Yes. Ripple handles the full preparation workflow: document intake, data extraction, customer rules, commodity code validation, CDS data structuring, pre-submission checking and exception routing. Experienced customs staff review and submit the final declaration through their existing system.

Does Ripple submit declarations to HMRC?

No. Ripple prepares submission-ready entries in your declaration software; your team reviews and submits them. How far Ripple goes towards submission is agreed per customer.

Does Ripple handle both imports and exports?

Yes. Every job is tagged with its direction first. Ripple then applies the right rules: 10-digit codes, valuation method, item-level origin and importer EORI on imports; 8-digit codes, exporter EORI and consignee details on exports.

Can Ripple validate commodity codes before declaration submission?

Yes. Ripple applies UK Global Tariff classification logic to every line item, checks code length and validity for the direction, checks procedure codes, origin evidence and preference claims, and flags mismatches before the entry reaches CDS. Genuine exceptions go to a named reviewer with a suggested answer.

What happens when data is missing from the documents?

Ripple emails the sender asking for the specific missing field, in a format they can answer in one line, and carries on when they reply. Your team does not have to chase.

Does Ripple integrate with ASM Sequoia and CargoWise?

Yes. Ripple writes structured, CDS-compliant data directly into ASM Sequoia and CargoWise, with two-way integration on CargoWise. The entry is created against the correct job, with header information, invoice rows and approved consolidation applied. Operators do not rekey data.

Does Ripple maintain a full audit trail for customs decisions?

Yes. Every entry includes the source documents used, the commodity code applied, the classification logic and interpretative rule referenced, any alternatives considered, and the human review outcome. Records are searchable, time-stamped and support HMRC’s four-year archiving requirement.

Can Ripple identify tariff impact when commodity codes or duty rules change?

Yes. Ripple applies classification logic centrally across all customers. When a tariff change affects a commodity code, the impact can be assessed across the relevant shipment population without reviewing entries one by one, and affected entries are flagged before submission. For large product catalogues, Ripple can bring the whole catalogue up to date in one project and keep it current, explaining what changed and why.

How long does it take to go live?

Ripple goes live in weeks. Each deployment is built around the client’s own customs rules, customer base and declaration workflow, with integration into the existing declaration system.

What evidence should freight forwarders require before choosing a customs automation vendor?

Look for live production results, not pilot outcomes. Confirm that the tool writes directly into your declaration system rather than generating XML for manual import. Test it on your most complex documents, not your simplest. Ask whether it needs a separate template per customer, and if so, who maintains them.

Further Reading

Ripple is not a licensed customs broker and this page is not customs advice. Rules, code structures and duty measures change frequently; check current requirements with HMRC before relying on anything here.

See Ripple Handle Your Customs Workflow

If your customs team is spending hours preparing declarations that should take minutes, or absorbing rework that should have been caught before submission, Ripple can show you what a different operating model looks like.

plus icon
CONNECT WITH US

Let Ripple Supercharge Your Operations  

high speed rail logistics
Contact Us