Dosto

What it is

An automation tool I built for Papillon Libros S.A., where I work. It replaces the manual data entry of commercial documents (invoices, credit notes, and delivery notes): it processes the PDFs, extracts the information, and runs automatic cross-checks of stock and returns.

What it does

• Converts invoice and credit-note PDFs into clean Excel sheets, mapping barcodes, ISBNs, and discounts row by row.
• Supports the formats of several large publishers (Penguin Random House, Santillana, Maipue, Heliasta), each with its own PDF structure.
• Compares delivery notes against credit notes to detect missing items, and even drafts the claim letter to the publisher automatically.

What I learned / solved

• Extracting data from PDFs that were never meant for it: each publisher builds its documents differently, so I used regular expressions (RegEx) and “fingerprints” like the tax ID (CUIT) to identify and parse each format.
• Pandas to cross-check inventories: grouping, consolidating, and comparing thousands of rows with outer joins to find discrepancies.
• Building a desktop interface usable by non-technical people: my coworkers open it with a double click, no terminal needed.
• Above all, solving a real problem that used to be done by hand and took hours.

Why I include it

It’s the project I’m most proud of: it came from a concrete need at work and is actually used today. It shows I can take code from a real-world problem all the way to a finished tool.