Admit the minimum
Freeze one field map for two agreed exports. No credentials, broad system copy or unrelated customer fields.
Founding-firm pilot · Australia
One queue for the exceptions your stack leaves open.
A read-only close-control workflow for ecommerce accounting and bookkeeping firms: compare two agreed exports across client stores, preserve unresolved states and produce a traceable review pack—without posting journals or replacing your current tools.
Contact is a request for fit review, not a pilot acceptance. Do not send credentials, files, client/store identities, customer records or authentic books.
The narrow workflow
See the remaining control work across stores.
Commerce, payment and accounting systems can each complete their own job while an identity, amount, fee, refund or timing difference remains. Settle Trace keeps that gap explicit rather than forcing a tidy result.
Preserve every state
Match declared references and amounts while missing, duplicate, stale, unsupported and unknown rows remain visible.
Export review evidence
Return per-store totals, one cross-client queue and an evidence pack bound to exact input and rule identities.
Mechanics demonstrated · buyer fit unverified
The prototype refuses a tidy total when the evidence is not tidy.
The example below is a deterministic synthetic fixture—not a customer result, connector test or accounting sign-off.
One fictional monthly portfolio view across four fictional clients.
Amount, missing refund reference, timing and unknown duplicate states.
Two ready, three review and one blocked—without forcing a clean close.
Known net difference, reported separately from unknown exposure.
A missing refund reference survives even where a store total agrees.
A duplicate settlement identity blocks safe monetary quantification.
Synthetic July 2026 fixture. No real firm, person, client, store, order, settlement or monetary event is represented.
Qualification before pitch
A fit only if one recurring gap survives the current stack.
The first conversation is meant to falsify the hypothesis quickly. “Already solved” is a useful answer; a generic wish for automation is not enough.
Possible fit
- Australian ecommerce accounting or bookkeeping firm
- At least five active ecommerce client stores
- One cross-system exception class recurs monthly
- Current tools and the remaining gap can be named
- A minimal read-only export boundary is feasible
Not this pilot
- Bookkeeping cleanup, journal posting or accounting advice
- Broad system replacement or every-provider support
- Direct write access, credentials or live database copies
- A problem the current stack already resolves
- Guaranteed savings, correctness or close completion
Evidence that matters
- One recent concrete occurrence and current workaround
- Frequency, store spread and review burden
- Safe input feasibility and decision authority
- A real A$500 accept, commit or decline decision
- No page view, compliment or generic interest substitutes
Provisional AUD pricing
Pay for one bounded result before software scale.
The price is a hypothesis, not accepted willingness to pay. The first evidence gate is one cleared A$500 pilot or two explicit A$500-per-month commitments from qualified firms.
First decision
Founding pilot
A$500AUD total · once if agreed · GST not charged
Included
One firm, one declared period, up to three stores, two export types, one field map, one run and one evidence pack.
Before any file
Written identity, scope, payment, safe transfer, privacy, IP, retention, deletion and acceptance terms.
Continuation
Only after a successful standardised pilot: intended Firm plan, A$500 total per month for up to ten stores while GST is not charged.
Not included
No bookkeeping, journal posting, tax/accounting advice, direct provider access, custom dashboard or promised outcome.
Fit review only
Tell us the recurring control gap—not the customer data.
Send only your work role, approximate active-store band, the two systems involved, a short description of the recurring gap and whether an A$500 decision is possible. Do not send files, credentials, client/store identities, customer records, transaction values or authentic books.
Claim boundary
A control layer, not a claim that the books are right.
A successful run means declared checks completed reproducibly. The firm retains professional judgment and decides how every exception is resolved.
Does it post journals or change source systems?
No. The founding-pilot workflow is read-only and export-first. It does not modify commerce, payment, accounting or bank records.
Is this accounting, tax, audit or legal advice?
No. It is a deterministic exception-and-evidence workflow. It does not certify completeness, correctness, compliance or fraud absence.
Which providers are supported?
No live provider compatibility is advertised. The synthetic example uses Shopify/Afterpay-shaped fields only. A qualified firm’s repeated gap selects two exact exports under a written field gate.
Is it generally available?
No. This is a founding-firm validation pilot. Product availability, marketplace approval and low-touch operation are not established.
How is AI used?
AI-assisted tools were used in product and document preparation. The pilot comparison uses frozen deterministic rules over two agreed exports; AI does not decide accounting treatment, approve a close, post to a source system or replace professional judgment. Buyer data is not sent to an AI provider in this pilot.