Automate Document Data Extraction with AI Presets: Set It Up Once, Use It Forever

If your team handles invoices, delivery notes, bank statements, or purchase orders on a regular basis, you already know the drill: open the document, find the numbers, copy them into Excel, repeat. It is tedious, error-prone, and eats hours that should go toward actual analysis.

AI-powered presets change that equation entirely. Instead of rebuilding your extraction logic from scratch every time a new document arrives, you define the rules once — and let the AI do the rest, every single time.

What is a "preset" and why does it matter?

A preset is a saved, reusable extraction template. You tell it what kind of document to expect (a scanned invoice, a bank statement, a supplier delivery note), which fields to pull (date, amount, item description, VAT, supplier name), and what the output should look like in Excel or CSV.

Once saved, the preset works like a one-click workflow. Drop in a new document — whether it is a clean PDF, a photo of a receipt, or a multi-page scanned file — and the AI maps the data to the right columns automatically. No reformatting, no hunting for values, no manual copy-paste.

Short answer: Presets let you extract structured data from repetitive documents — PDFs, images, scanned files — into Excel with a single click, using AI that remembers exactly what you need and how you need it.

Tablola offers a library of ready-made presets for the most common document types so you do not have to build from zero. You can start with a template like Invoice to Excel or Bank Statement to Excel or CSV and customise the columns to match your own naming conventions and structure.

Three document workflows that save the most time

1. Invoices and purchase orders

Finance teams process dozens — sometimes hundreds — of invoices each month. Every invoice contains the same core fields: vendor, date, line items, unit prices, totals, tax. A preset captures all of that in one pass. The Fatura Verisini Excel'e Aktarma preset and its English counterpart handle both digital and scanned invoice PDFs, extracting line-level detail into a clean, sortable spreadsheet.

Purchase orders follow the same logic. Once you configure the Purchase Order to Excel preset, every new PO from every supplier feeds into the same master table without any manual work.

2. Bank statements and financial exports

Reconciling accounts manually from multi-page bank statement PDFs is one of the highest-friction tasks in any accounting workflow. A single preset can parse transaction dates, descriptions, debits, credits, and running balances across hundreds of rows — in seconds. The output is a properly structured CSV or Excel file ready for pivot tables or accounting software import.

3. Delivery notes and goods receipts

Warehouse and logistics teams deal with delivery notes from multiple carriers and suppliers, each with a slightly different layout. AI-based extraction handles layout variation far better than rigid template-matching tools. You define the fields you need — item code, quantity, batch number, delivery date — and the model figures out where they live on the page, even when the format shifts. The Delivery Note to Excel preset is a solid starting point for this workflow.

How to set up your own preset in Tablola

1. Choose a base template. Browse the preset library and pick the document type closest to your use case. Starting from an existing template saves significant setup time.
2. Upload a sample document. Tablola uses the sample to identify the available data fields and preview the extraction output.
3. Map and label the columns. Rename columns to match your internal conventions (e.g. "Inv. No." instead of "Invoice Number"), reorder them, and remove any fields you do not need.
4. Save the preset. Give it a clear name tied to the document type and supplier or source if relevant.
5. Run it on new documents. From that point on, any document of the same type can be processed with one click. Results land in Excel or CSV, ready to use.

If you handle batches — say, 50 invoices from the same supplier at month-end — the Merge Multiple Documents into One Table preset combines all of them into a single consolidated spreadsheet in one operation.

What makes AI extraction different from older tools

Older PDF-to-Excel converters worked by position: they expected column A to always start at a fixed pixel coordinate. The moment a supplier changed their invoice template, the extraction broke.

AI extraction understands meaning, not just position. It recognises that "Total excl. VAT", "Net amount", and "Subtotal" all refer to the same concept, and it adapts to layout changes automatically. This makes presets genuinely durable — they keep working even when the source documents evolve.

Combined with the ability to extract from images and photos (not just clean PDFs), this means even a blurry photo of a printed receipt can be turned into structured data. The Receipt Photos to Excel preset handles exactly this case for expense reporting and petty cash reconciliation.

Frequently asked questions

Do I need technical skills to create a preset?

No. Presets are built through a visual interface — you upload a sample document, select the fields you want, and save. There is no coding or formula writing involved. Most users have a working preset within a few minutes of starting.

Can one preset handle documents from different suppliers with different layouts?

In most cases, yes. Because the extraction is AI-driven rather than position-based, it adapts to moderate layout variation automatically. If two suppliers use very different structures, it is usually cleaner to maintain a separate preset for each — setup takes only a few minutes per supplier.

What file types can I use as inputs?

Tablola accepts standard PDF files (both digital and scanned), JPEG and PNG images, and multi-page documents. This covers the vast majority of real-world document sources: email attachments, scanned archives, photos taken on a phone, and exported reports.

Is the output always an Excel file?

You can choose between Excel (.xlsx) and CSV depending on your downstream workflow. CSV is useful when you are importing directly into accounting software or a database. Excel is better when you plan to do further analysis, filtering, or formatting within the spreadsheet itself.