We're building the bridge between messy paper and clean data.

Our mission

Spreadsheets run the world. Bookkeepers, analysts, ops leads, lab managers — they all live inside Excel. But getting data into the spreadsheet is still mostly manual.

Parsebay removes that step. Drop in a PDF and the data lands in a real spreadsheet: edit it, put formulas in it, keep feeding it new documents every month, and download Excel, CSV or JSON whenever you need to.

Why it exists

Most extraction tools want teaching first: draw boxes around the fields, build a template per supplier, then maintain it as layouts drift. That works when a hundred documents look identical. It falls apart when they do not — which is most of the time for a business taking invoices from forty different companies, each with its own idea of where the total goes.

Parsebay was built the other way round. It reads the page the way a person would and reports what is actually on it, rather than matching it against a shape somebody defined in advance. That is why there is nothing to configure, and why a document you have never seen before behaves the same as one you process every month.

Who it's for

Anyone who retypes numbers off a document into a spreadsheet. Bookkeepers working through a month of supplier invoices. Ops and admin staff with delivery notes, purchase orders and receipts. Analysts pulling figures out of statements that only ever arrive as PDFs.

There is no template to configure and no field mapping to set up, so it does not matter whether every document looks the same. If a page has numbers laid out in rows and columns, Parsebay reads it — including scans and photos, where the text cannot be selected at all.

Who builds it

One person. Not a team — one developer, who wrote the extraction pipeline, the spreadsheet you edit in, and this page.

The belief behind it is simple: if a human can read it, software should be able to too. Questions or feedback? Get in touch — it comes to me.