Switching from desktop software or a spreadsheet should be painless.
CSV import
The guided import wizard (Import in the sidebar) lets you upload a CSV export from Moneylender Pro, NoteSmith, or a generic spreadsheet. You pick your file, NoteHarbor auto-maps the columns it recognizes and you finish mapping the rest, preview exactly what will be created, then import — loans, borrowers, and properties in one pass. CSV import is always free.
Other file types (TXT, Word, OpenDocument)
Besides CSV and PDF, the wizard also accepts plain text (.txt), Word (.docx and legacy .doc), and OpenDocument text (.odt). It reads whatever the file holds and drops it into the same auto-map then finish mapping flow:
- A table of loans in a Word/OpenDocument file, or delimited columns in a .txt, is turned into rows automatically — free, no AI, just like a CSV.
- An unstructured Word or text document (prose, not a table) is read with AI — the paid Deep Analysis add-on, same as PDF.
- A legacy .doc that can't be read automatically gets a friendly prompt to re-save it as .docx or export a CSV. Nothing is ever imported as garbage.
Whatever the source, you always auto-map, finish mapping the rest, preview, and confirm before a single loan is created. A NoteSmith Payment History report is recognized whether it's a .csv or a .txt.
Fresh vs. current-state
A row with only origination terms (principal, rate, term, start date) is created brand-new and the engine projects its full schedule. A row that also carries a current balance plus a temporal marker — a next-due date, payments made/remaining, or a last-payment date — imports at its live state: seeded balance, next-due date and payments remaining, with the go-forward schedule built from that balance at the contractual payment. The preview labels every loan Current state or Fresh and flags oddities (e.g. a balance above the original principal) as warnings.
Switching from NoteSmith
NoteSmith exports plain CSV, and its Pro custom report generator (Report Manager) exports Excel/CSV with the fields you choose — there's no fixed layout, so the wizard auto-detects NoteSmith's usual column names and you correct anything before importing. Build a report with one row per loan, export it as CSV, and include: account/note number, payor name, property address, original amount, interest rate, term, first-payment date, unpaid/current balance, next payment due date (or payments made/remaining), the P&I payment, and — if handy — maturity date, escrow balance, interest-paid-to-date, and lienholder/investor.
Statement PDFs (terms interview + payment history)
Have one loan's servicer statement or payment-history printout as a PDF (e.g. a NoteSmith "Payment History" report you printed to PDF)? Use Import a statement PDF on the Import page. NoteHarbor reads the PDF's text and:
- Extracts the loan terms — principal, rate, dates, payment, current balance — each tagged with how confidently it was read. High-confidence values are pre-filled; anything read shakily (or not at all) is highlighted and you confirm or fill it before anything is created. We read what we could — you confirm the rest.
- Reads the payment ledger and verifies it the same way the CSV history import does: every payment's printed interest/principal split and running balance must reconcile to the cent, ending exactly at the current balance. A ledger that doesn't tie is flagged for review, never imported silently — you can still import the loan without its history.
- Creates the loan with its history attached and saves the source PDF to the loan's Documents (as Payment history), so the original artifact stays with the loan.
The built-in reader is free. With the paid Deep Analysis add-on, AI reads messy or unusual statement layouts far more reliably. Scanned image PDFs can't be read — you'll get an honest message to enter the loan manually or upload a text-based PDF or CSV; nothing is ever guessed from pixels.
Portfolio PDFs (Deep Analysis)
For a multi-loan portfolio list or schedule in PDF, the regular PDF option reads the loans into reviewable rows via AI. This uses the paid Deep Analysis add-on — CSV is free, so prefer it when you can export one. As with statement PDFs, we read the PDF's text, so scanned image PDFs can't be read — you'll get an honest message pointing you to a CSV export or a text-based PDF rather than a guess.
A current-state-imported loan services forward normally; while it has no payment history, editing its terms or past payments is held so its balance is never rebuilt from the wrong starting point. Need a bulk multi-loan history migration? Send your export to support.