Duty, Tax & Financial Audit
The three working screens the Customs Declarant and Customs Manager use to calculate landed cost, reconcile duty and tariff charges, and validate trade documents.
What this group does
Duty, Tax & Financial Audit closes the loop on a declaration's financial side: calculating what a shipment actually costs once freight, duty, tax and fees are included, keeping a ledger of duty and tariff charges for finance reconciliation, and using AI assistance to catch mismatches between trade documents and declarations before they become audit findings.
Screen preview
A quick look at the screen. Use the toggle in the corner to switch between the dark and light themes.
Duty and tariff ledger with landed cost
1. Landed Cost Calculator
Purpose: Calculate total landed cost — product, freight, duty, tax and fees — per shipment.
Rolls every cost component of a shipment into one figure, so the true delivered cost of goods is known before pricing or margin decisions are made.
2. Customs Duty & Tariff Ledger
Purpose: Ledger of duty and tariff charges paid per declaration, for finance reconciliation and audit.
Keeps a running record of what was paid in duty and tariffs on each declaration, giving Finance and the Customs Manager a clean trail for reconciliation and for responding to an audit.
3. Doc Matcher & Linter (AI)
Purpose: AI-assisted matching and validation of trade documents against declarations, flagging mismatches.
Compares the supporting trade documents attached to a shipment against its declaration and flags mismatches automatically, so a discrepancy is caught by the Compliance Officer before a customs authority finds it first.
Who can do what
| Step | Typical role |
|---|---|
| Calculate landed cost for a shipment | Customs Declarant / Broker Liaison, Customs Manager |
| Record or reconcile a duty/tariff charge | Customs Declarant / Broker Liaison, Customs Manager |
| Approve an audit finding | Compliance Officer, Customs Manager |
| Review an AI document match | Compliance Officer, Customs Declarant / Broker Liaison |
AI Assistant — Landed Cost Calculator
The Landed Cost Calculator screen ships with an embedded AI Assistant — the same right-edge panel used across TotalApp. Open it from the vertical AI Assistant tab on the right of the screen; it understands what is currently on your screen and answers in Markdown, matching the language you write in (Turkish or English).
Because an unexpected duty/VAT combination can quietly erode margin on a shipment, the assistant is tuned for margin-impact review:
- High margin erosion — flag calculations where landed cost has eaten unusually deep into the expected margin.
- Unusual duty/VAT combinations — surface calculations whose duty and VAT combination looks inconsistent with similar shipments.
- Cost-breakdown explanation — walk through what is driving a specific calculation's total landed cost.
- What-if guidance — discuss how a different Incoterm or origin might change the landed cost outcome.
Tick one or more rows in the table to scope the conversation, then use Analyze selected or Summarize selected. Any answer can be saved via Save as Report (it lands in My Reports), and Add Knowledge lets you attach reference documents to ground the reply.
Engine-aware
The assistant respects your Settings → Agentic writer engine: Ollama / local-LLM / in-browser models run entirely on your machine, while API routes through the server — the panel never leaks your data to a server the selected engine can’t reach.
AI Assistant — Customs Duty & Tariff Ledger
The Customs Duty & Tariff Ledger screen ships with an embedded AI Assistant — the same right-edge panel used across TotalApp. Open it from the vertical AI Assistant tab on the right of the screen; it understands what is currently on your screen and answers in Markdown, matching the language you write in (Turkish or English).
Because an entry stuck under appeal or an uncovered escrow balance is real, uncollected liability, the assistant is tuned for tax-liability review:
- Under-appeal / refund-claimed entries — surface ledger entries currently in dispute or awaiting a refund decision.
- Escrow coverage gaps — flag entries where the escrowed amount does not fully cover the assessed liability.
- Aging review — highlight entries that have sat unresolved the longest.
- Liability summary — summarize total outstanding duty exposure across a selected set of entries.
Tick one or more rows in the table to scope the conversation, then use Analyze selected or Summarize selected. Any answer can be saved via Save as Report (it lands in My Reports), and Add Knowledge lets you attach reference documents to ground the reply.
Engine-aware
The assistant respects your Settings → Agentic writer engine: Ollama / local-LLM / in-browser models run entirely on your machine, while API routes through the server — the panel never leaks your data to a server the selected engine can’t reach.
AI Assistant — Doc Matcher & Linter
The Doc Matcher & Linter screen ships with an embedded AI Assistant — the same right-edge panel used across TotalApp. Open it from the vertical AI Assistant tab on the right of the screen; it understands what is currently on your screen and answers in Markdown, matching the language you write in (Turkish or English).
This chat assistant is a separate, complementary feature from the screen's existing structured two-stage Linter/Critic audit — the audit runs a fixed cross-document consistency check and returns a structured discrepancy report, while this assistant is an open-ended conversation you can steer, ask follow-up questions in, and use to explore results the audit already surfaced. It is tuned for cross-document integrity discussion:
Because this is a conversation rather than a fixed audit run, use it to dig into a discrepancy the Linter/Critic audit already flagged, or to ask a question the structured report doesn't answer directly.
- Critical discrepancies — explain the highest-severity mismatches found across a document bundle in plain language.
- High discrepancy scores — discuss which document pairs or bundles carry the highest overall mismatch score and why.
- Root-cause discussion — walk through what likely caused a specific field-level mismatch (e.g. a quantity or value that doesn't line up between documents).
- Resolution guidance — suggest which document is most likely to be the error source and what to check next.
Tick one or more rows in the table to scope the conversation, then use Analyze selected or Summarize selected. Any answer can be saved via Save as Report (it lands in My Reports), and Add Knowledge lets you attach reference documents to ground the reply.
Engine-aware
The assistant respects your Settings → Agentic writer engine: Ollama / local-LLM / in-browser models run entirely on your machine, while API routes through the server — the panel never leaks your data to a server the selected engine can’t reach.