Bank statement to Pennylane
Turn any bank's PDF relevé bancaire into a clean CSV of dated, signed transactions you can bring into Pennylane — no retyping, no line-by-line entry. Any bank, scanned or digital, balance-validated before you import.
Any bank · scanned or digital · balance-validated · CSV / Excel
Get PDF relevés bancaires into Pennylane without retyping
Pennylane centralises a company's accounting around live bank connections and its own transaction aggregation, which is excellent when every account is connected and you only care about going forward. FlowParse is not a replacement for Pennylane's feed — it is the extraction layer for what that feed can't reach: a bank Pennylane doesn't aggregate, a foreign or business account outside the connection, or the historical months from before you onboarded. For those, a PDF relevé bancaire is the record, and FlowParse turns it into a clean CSV of dated, signed écritures you can bring into Pennylane.
FlowParse reads every transaction off the relevé bancaire — the date, the libellé and a signed amount — and builds a clean CSV for Pennylane instead of you keying each line. The same upload also produces a full Excel workbook with every source column preserved, so you keep detail the Pennylane transaction view doesn't show.
Because extraction is AI-based rather than template-based, it reads any bank's layout on the first try — and every file is balance-validated so the figures you import into Pennylane are complete and correct, not a hopeful copy-paste.
Two ways in
In the file
From PDF to Pennylane in three steps
1 · Upload the PDF
Drop your PDF bank statement — digital or scanned, any bank, any number of pages.
2 · AI builds the file
FlowParse extracts every transaction, signs the amounts, validates balances and writes a CSV.
3 · Import into Pennylane
Bring the file into the matching account in Pennylane and reconcile — no manual entry.
How to bring relevés bancaires into Pennylane
Convert first: upload the PDF relevé bancaire to FlowParse, review the editable preview, and download the CSV. Every row is a transaction with a date, a description and a signed amount, ready to use.
Then import: use the CSV to import the transactions into Pennylane, mapping the Date, Description and Amount columns once. Where Pennylane offers a live bank connection for your bank, use it for the day-to-day feed and use FlowParse for the banks it can't reach and for historical catch-up.
Have a supplier bill to record too? FlowParse also maps an invoice to a clean line for your books, so the same tool covers both sides of your bookkeeping — statements and invoices — into Pennylane.
What lands in the CSV
| Field | From the statement |
|---|---|
| Date | Transaction date, normalised |
| Description | Payee / libellé (extra columns kept in Excel) |
| Amount | Signed — money out negative, money in positive |
Works with any bank's PDF
Because extraction is AI-based rather than template-based, FlowParse reads layouts it has never seen — ideal for accounts Pennylane's feeds don't cover and for historical months:
Reconciling a whole year? Use Smart Merge to consolidate first, then export once.
Why convert instead of typing it in
Manual entry is where small bookkeeping errors are born: a transposed figure, a skipped line on a long relevé bancaire, a wrong sign on a refund. Each surfaces later as a reconciliation that won't balance and an hour spent hunting the difference. Converting removes the typing entirely — and the balance check proves nothing was dropped before it reaches Pennylane.
For an expert-comptable handling a rattrapage — or a business catching up its own books — that is the difference between an evening of saisie and a few minutes of upload-and-review. And because FlowParse keeps the full detail in Excel too, you never lose a column Pennylane doesn't store.
How Pennylane handles bank data — and where this fits
There is a coverage gap and a history gap, and both are exactly where a converter fits. Not every French bank, néobanque or foreign account is aggregated, and a live connection only starts from the day it is linked, so the back-months an expert-comptable needs for a rattrapage simply aren't in the feed. FlowParse reads each relevé — date, libellé and a signed amount — and rebuilds the period as clean rows, balance-checked, ready for Pennylane.
There is also a quality angle. A live feed sometimes imports terse, codified descriptions that are hard to categorise against TVA; a PDF relevé bancaire usually carries the fuller narrative. Because FlowParse preserves the complete libellé (and keeps every other column in the matching Excel export), you get richer detail to work with, not less — and because the figures are balance-validated before import, you are not trading convenience for correctness.
The practical workflow is simple: keep Pennylane's live connection for your main bank, and use FlowParse for everything it can't reach — a second account, a card from a smaller issuer, a foreign account, or the back-months you need to get caught up. One tool, any bank, no retyping, and a clean CSV at the end that Pennylane takes cleanly.
Validated before it reaches Pennylane
A bad import means a reconciliation that never balances. Before the file is built, FlowParse checks the data so what lands in Pennylane is complete and correct:
See the validation engine and reconciliation tools.
Accuracy, security and a real catch-up
Accuracy is the whole point of converting rather than typing, so it is worth knowing how it is achieved. FlowParse reads each relevé bancaire by meaning rather than fixed coordinates, which is why an unfamiliar bank layout is handled on the first try and a redesigned statement keeps working. Scanned or photographed statements are read with OCR first, then structured into transactions, and every figure is checked against the statement's own balance before you ever import into Pennylane. Any line the engine is unsure about is flagged for a quick look rather than silently guessed.
Because bank statements are sensitive, data handling matters as much as extraction. FlowParse processes documents in EU data centres, deletes the original PDF as soon as extraction completes, stores only the extracted data (encrypted, and deletable on demand), and never uses your documents to train models — the full posture is on the security page. For an expert-comptable handling client files, that means you can run a relevé bancaire through without worrying where it ends up.
Picture a typical catch-up: a client joins Pennylane midway through the year with nine months of activity across a main account and a second account the feed doesn't support. Instead of an evening of manual entry, each PDF is converted, the balance check confirms every period is complete, and clean transactions import ready to categorise against TVA — the whole back-year reconciled in an afternoon rather than a week.
And because the same upload also yields a full Excel workbook, nothing is lost: Pennylane holds the transactions it needs, while the workbook keeps every original column for your records, audit trail or expert-comptable. One conversion, both outputs, any bank — the honest, complete way to get a relevé bancaire into Pennylane.
One upload, every format
The same extraction powers every export — so if you also run QuickBooks, Quicken, Xero or just need a spreadsheet, take the same statement to the format you need without re-uploading.
Take the same data to Excel, QuickBooks or Xero.
| Format | Best for |
|---|---|
| CSV | Importing this statement into Pennylane |
| Excel (.xlsx) | A full workbook with every source column for your records |
| CSV | A simple, editable table for any tool or import wizard |
| OFX / .QBO / .QFX | Bank-feed files for QuickBooks, Quicken and OFX apps |
| Xero / Sage / Wave | Tailored CSV layouts for other accounting software |
Who imports statements into Pennylane this way
Anyone who reconciles a Pennylanebank or card account from statements rather than a live feed — because the feed doesn't cover the account, hasn't been set up yet, or the months they need predate it. A few of the most common:
Expert-comptables
Onboard a client's back-months into Pennylane without keying a year of relevés.
TPE / PME
Bring in a second or foreign account Pennylane doesn't aggregate.
Startups
Backfill pre-Pennylane history from PDF relevés before the first close.
Freelances
Import statements from a bank the feed can't reach.
Cabinets comptables
Standardise mixed-bank PDFs into one Pennylane-ready CSV per client.
Finance teams
Reconcile accounts outside the aggregation against clean, signed data.
Any bank → one Pennylane-ready file
Upload statements from every account you reconcile — FlowParse turns each into a clean file Pennylane imports.
Tips for a clean import into Pennylane
A statement import goes smoothly when the data underneath is complete and correctly signed. A few habits make every Pennylane import reconcile on the first pass:
- Convert the full statement period rather than a partial export, so the opening and closing balances are present for the validation check to confirm nothing is missing.
- Import each statement period once — or use the OFX file where Pennylane supports it, since it de-duplicates by a per-transaction ID if you re-import.
- Match the Pennylane account's currency to the statement you're importing; FlowParse preserves the amounts and signs exactly as the bank reported them.
- Review the fields the validation report flags in the editable preview before you download — clean statements pass automatically, so you only check the exceptions.
- Reconciling a whole year? Consolidate the statements with Smart Merge first, then import a single clean, balance-checked set into Pennylane.
Frequently asked questions
How do I import a bank statement into Pennylane?
Convert the PDF relevé bancaire with FlowParse to a clean CSV of dated, signed transactions, then import the CSV into Pennylane and map the Date, Description and Amount columns once. For banks Pennylane connects to live, use its feed; use FlowParse for unsupported banks and historical months.
Does Pennylane import a CSV bank statement?
FlowParse is not a replacement for Pennylane's feed — it is the extraction layer for what that feed can't reach: a bank Pennylane doesn't aggregate, a foreign or business account outside the connection, or the historical months from before you onboarded. FlowParse gives you a clean, correctly-signed CSV plus a full Excel workbook, so the transaction data is ready however you bring it into Pennylane.
Can I convert credit-card statements for Pennylane?
Yes. Credit-card PDF statements convert the same way — each line becomes a dated, signed transaction — so you can record card spending in Pennylane alongside bank activity.
What about multi-page statements?
Long, multi-page relevés bancaires are stitched into one continuous list and weak pages are retried automatically, so nothing is dropped at page breaks before the CSV is built.
Can I import historical or missing relevés bancaires?
Yes. Convert any PDF you have — including older months your bank only provides as a relevé bancaire PDF — and bring them into Pennylane to backfill the period before you reconcile or file.
Will I lose any detail from the statement?
No. The CSV keeps Date, Description and Amount, and the matching Excel export preserves every original column 1:1 — so any reference, balance or category column on the relevé bancaire is retained for your records.
Is it free to convert a statement for Pennylane?
You can try conversion within the monthly page allowance — upload a relevé bancaire and preview the extracted transactions first. Accounting exports are part of the paid plans for ongoing volume.
Can I convert statements from any bank for Pennylane?
Yes. FlowParse uses AI extraction rather than per-bank templates, so PDF statements from UK banks (Barclays, HSBC, Lloyds, NatWest, Monzo, Starling), US banks (Chase, Bank of America, Wells Fargo, Capital One), and EU and neobanks (Wise, N26, Revolut) all convert to a Pennylane-ready file. A layout the tool has never seen is read correctly on the first try.
Are scanned or photographed statements supported?
Yes. Image-only PDFs run through OCR first, then the AI structures the recognised text into transactions before building the file for Pennylane. Any low-confidence field is flagged so you can check it before importing.
How are debits and credits handled?
They are normalised into a single signed amount — money out negative, money in positive — with the correct transaction type, so the imported balance reconciles against your account.
Is the conversion accurate?
FlowParse reaches around 98% field-level accuracy on standard statement formats, and every file is balance-validated (opening balance + transactions = closing balance) with per-field confidence scores you can review before importing into Pennylane.
Can I review the transactions before importing?
Yes. An editable preview lets you correct any value, and the validation report flags duplicates, low-confidence fields and balance breaks before you download the file — so nothing wrong reaches your books.
Do I also get other formats from the same upload?
Yes. One upload can export Excel, CSV, OFX and the Intuit-tagged .QBO and .QFX for QuickBooks and Quicken, plus Xero, Sage, Zoho Books, NetSuite and MYOB — so multi-tool firms convert once and export everywhere.
Where are my bank statements processed and stored?
FlowParse processes documents in EU data centres, deletes the original PDF as soon as extraction completes, stores only the extracted data (encrypted, and deletable on demand), and never uses your documents to train models. So converting a statement for Pennylane is safe, not a privacy trade-off — the full posture is on the security page.
What happens with multi-page or very long statements?
They're stitched into one continuous list and any weak page is retried automatically, so nothing is dropped at a page break. The balance check then confirms the whole statement — first page to last — was captured before the file is built.
Can I automate converting statements for Pennylane at volume?
Yes. The same conversion runs over the bank statement API: send a PDF and receive structured JSON or a ready-to-import file per page, with the balance validation built in. That lets high-volume teams turn statement intake for Pennylane into a pipeline step instead of a manual task.
Do I have to enter any transactions by hand?
No. FlowParse extracts every line from the PDF and writes them into the file, so Pennylane loads the transactions in one import instead of you typing each one. You only review the few fields the validation report flags.