Getting started with LedgerSync
For buyers, AP clerks and tenant admins: the order that makes everything match, where to click, and what each status means.
You send supplier invoices. LedgerSync reads them and checks them against the invoices already in your ERP. You only look at the ones that do not line up. The same walkthrough, with live examples, is under Help in the app after you log in.
What to expect
- Fully matched — the extracted invoice lined up with an ERP row on every field that could be compared. You can move on.
- Partially matched — close, but a field disagrees or is missing. That is the review queue, not a failure.
- Not matched — no ERP row was found. Import the catalog (or that row), then Match again.
- You do not confirm column mappings, re-extract or hand-reconcile as a ritual. Matching runs by itself after upload.
The order that matters
ERP catalog first, then PDFs
- Log in with the email and password from your invitation.
- ERP Records → pick the format pill (CSV / XLSX / XML / JSON / URL) → drop your ERP export.
- Documents → drop invoice PDFs or images into the dashed box — one file, many files, or one bulk PDF.
- Wait until the status reads PROCESSED.
- Open a row, or go to Reconciliation and filter by status.
Where to click
| Go here | Click / drop | Why |
|---|---|---|
| Login | Email + password → Log in | Opens your organisation. No public signup. |
| ERP Records | Format pill → drop file | Loads the invoices your ERP already knows. |
| ERP Records → “If a match looks wrong” | Only if a total or supplier bound to the wrong column | Optional fix. Skip it on a first run. |
| Documents | Dashed box: drop PDFs | Upload before you look at Reconciliation. |
| Documents table | Click a filename | Extraction on the left, match on the right. |
| Document detail | Download original file | Compare the PDF to what was read. |
| Document detail | Refresh from file | Re-read the PDF (older documents, or a failed extraction). |
| Document detail | Match again | Re-score against ERP without re-reading the PDF. |
| Document detail | Try again | Only when the status is FAILED. |
| Reconciliation | Filter Fully / Partially / Not matched | Work the review queue. |
| Reconciliation | Checkbox → Approve / Reject | Human confirmation on a partial match. Bulk actions need a reason. |
| Overview | Stat cards | Counts: documents, fully matched, needs review. |
| Settings → Security | 2FA | Optional two-factor authentication for your own account. |
| Settings → Team | Invite users | Tenant admin only. |
Viewers can read lists and detail. They cannot upload, import ERP records or approve.
What to upload
ERP export (first). Export open or posted supplier invoices from your ERP as CSV, XLSX, XML or JSON, or paste a URL. Keep your real column names (VendName, Invoice No., GrossAmt, Belegnummer, …) — the binding is inferred on import. Include invoice number, supplier, date, amounts and currency when you have them; extra columns become attributes and can still score. For an ongoing feed, see Settings → Integrations (SFTP or database).
Invoices (second). PDF or image, up to 50 MB each. One invoice per file, many files at once, or one bulk PDF. A pack is split wherever a page carries an invoice-number label (Invoice No., RECHNUNG, Belegnummer, …); cover pages without that label are skipped.
What you should see
- Documents list: filename, PROCESSED, supplier, invoice number, total, match badge.
- Document detail — Extraction: header (supplier, invoice number, date, subtotal / tax / total); line items (SKU if present, description, extra columns such as colour, quantity, unit price, amount); “Other extracted fields” (freight, ILN, payment terms, buyer, …); the raw extraction JSON, collapsed.
- Document detail — Reconciliation: match status, score and ERP record key; a field scores table with field, score, extracted value and ERP value.
- Reconciliation: fully matched at 100% on a clean catalog; partials in the amber card.
Specific cases
| Situation | What to do |
|---|---|
| First time / new organisation | Import ERP, then upload a small sample — not the whole archive. |
| Multi-invoice PDF | Drop the pack on Documents. Expect one row per invoice, not one row per file. |
| German or mixed labels (RECHNUNG, Warenwert, Farbe) | Nothing to configure. Extraction maps those. |
| ERP columns look ugly | Import anyway. Open “If a match looks wrong” only if a slot bound to the wrong header. |
| Match is partial and you agree with the PDF | Reconciliation → Approve (add a note on bulk actions). |
| Match is partial and the ERP is right | Fix the ERP export or the bound column, then Match again. |
| Not matched | Confirm the invoice exists in ERP Records (search the invoice number). Import that row, Match again. |
| Extraction looks wrong against the PDF | Refresh from file, then Match again. |
| FAILED status | Try again. If it keeps failing, check the original file is a readable PDF or image. |
| Need a second opinion | Open a ticket against the document. |
How matching works, in short
- You import ERP rows. Each row’s attributes are stored and the columns for invoice number, supplier, totals, date and currency are inferred.
- You upload a PDF. A background worker reads it into structured JSON — header, line items and extra fields.
- ERP rows are looked up by normalised invoice number and supplier. If nothing hits, the mapping is re-inferred from sample rows and the lookup runs again.
- The candidate is scored field by field: invoice number, supplier, total, date, subtotal, tax, currency, plus custom attributes and line items.
- The result is Fully, Partially or Not matched. You only act on Partially and Not.
The whole loop is import ERP → upload invoice → automatic extract → automatic match → review exceptions. The details, with real engine output, are in how an invoice is scored against an ERP record.
Questions we get
Every document extracts, but every match says “Not matched”. Why?
The ERP catalog is empty, or the invoice is not in the export you imported. Import the ERP file first, then search ERP Records for that invoice number. Once the row is there, open the document and click Match again.
What does “Partially matched” mean?
A record was found, but at least one field disagrees (amount, date, tax, a line item) or is missing on the ERP side. Open Field scores to see the extracted value next to the ERP value. If the PDF is right, Approve. If the ERP is right, fix the export or the bound column and click Match again.
A document shows FAILED. What now?
Click Try again. If it keeps failing, check that the original file opens as a readable PDF or image on your own machine; a scan with no readable content or a corrupted file is the usual cause. If a readable file still fails, contact us.
My upload was rejected with a size error.
Files are limited to 50 MB each and must be a PDF or an image. Split a large pack into smaller PDFs, or export the document again at a lower resolution.
I uploaded one PDF with many invoices. Why do I see many rows?
That is expected. A pack is split into one document per invoice wherever a page carries an invoice-number label such as Invoice No., RECHNUNG or Belegnummer. Cover pages without that label are skipped.
The invoices are in German, or mix German and English labels. Do I need to configure anything?
No. Labels such as RECHNUNG, Belegnummer, Warenwert, MwSt and Farbe are mapped at extraction time, and ERP column names such as Belegnummer or Rechnungsbetrag are recognised on import.
A colleague cannot upload or approve anything.
Their role is Viewer. Viewers can read lists and detail pages but cannot upload, import ERP records or review matches. A tenant admin can change the role under Settings → Team.
Login fails. Can I sign up myself?
There is no public signup. Organisations are created by the LedgerSync team, and users are invited by their organisation’s admin. Use the invitation email you received; if you do not have a workspace yet, contact us.
Ready to see it on your own invoices?