Productivity6 min read

Why I Built DATANEM: From Invoice Images to Structured Data

Discover why I built DATANEM to turn invoices and documents into structured data, with reusable extraction fields and exports to Excel, CSV and more.

Why I Built DATANEM: From Invoice Images to Structured Data

Datanem

Verified Tool

Instantly convert your documents into structured data

DATANEM is a document extraction tool that converts PDFs, scans, images and Word documents into structured datasets. Users can customise fields, organise information into tables and export data to spreadsheets or databases. Built to reduce manual data entry, it offers a free document trial without an account.

I built it because I had hundreds of invoice images to work through for grant reporting. Copying the information into a spreadsheet by hand would have taken weeks. I could see the details I needed, but getting them into a usable dataset was another job entirely.

I work in data and IT for a non-profit, and I am the founder and solo developer behind DATANEM. The product started with that practical problem: how do you turn information spread across documents into something you can actually use?

Table of contents

  • The work before the work

  • How DATANEM turns documents into data

  • A practical invoice example

  • Manual data entry and document extraction compared

  • Where the extracted data goes

  • What it costs to get started

  • How to test it on your own work

The work before the work

A finished report rarely shows how much preparation went into it. Before you can compare spending, check records or answer a question, the information needs to be in a consistent format.

An invoice might arrive as a PDF, a scan or a photograph. Different suppliers arrange the same information in different ways. The date is at the top of one document and near the bottom of another. A total might sit beside several other figures that look just as relevant.

Someone then has to find the right details and put them into the right columns. Repeat that across a pile of documents and the preparation becomes a substantial part of the task.

I wanted to choose the information I needed and build a dataset from the documents containing it. That became DATANEM.

You put documents in and get databases out.

Alexander Finlay Green, founder of DATANEM

How DATANEM turns documents into data

DATANEM demo showing a Suppliers lookup table connected to an Invoices table in the database designer.

The process starts with a document. You can upload a PDF, Word document, scan or photograph. Scans and photographs go through optical character recognition so the written information can be read.

For the anonymous trial, DATANEM proposes columns from the document itself and returns a table. You do not have to design a database before seeing whether the result is useful.

With an account, you can customise the fields you want to extract. That might mean renaming a column, removing something you do not need or adding a field relevant to your report.

Saved designs are available on paid plans for recurring work. Once you have a structure that suits a task, you can reuse it for later batches. DATANEM supports batches of up to 250 files, with a 10 MB limit per file.

The point is to make the structure fit the question you are trying to answer. A spending report and a supplier register might use the same invoices but need different columns.

A practical invoice example

DATANEM demo results showing six sample invoices organised into separate Suppliers and Invoices tables.

Suppose you need a spreadsheet showing the supplier name, invoice number, invoice date, net amount, VAT and total for a set of invoices.

With manual entry, you open each file, find each value and type or paste it into the sheet. You also need to keep the columns consistent as you move between different layouts.

With DATANEM, those fields become the extraction structure. The intended result is a table where the supplier name sits in the same column for every document, regardless of where it appeared on the original page.

You can then review the output, check figures against the source documents and export the dataset. This checking step matters, especially when information will be used for payments, accounting or grant reporting.

The same approach can be useful for delivery notes, purchase orders and timesheets. The fields change with the task, but the underlying need is similar: collect specific information from documents in a consistent form.

Manual data entry and document extraction compared

This comparison describes the workflow. It is not a speed benchmark or a claim that extraction will be correct on every document.

Task

Manual entry

Using DATANEM

Set the structure

Create spreadsheet columns.

Try proposed columns or customise fields with an account.

Collect values

Read and copy each value.

Extract values from uploaded documents.

Repeat the task

Continue entering the next set of documents.

Reuse a saved design on a supported paid plan.

Check the result

Check transcription against the originals.

Review extracted values against the originals.

Use the dataset

Work in the completed spreadsheet.

Export the data into a spreadsheet or another workflow.

Where the extracted data goes

I wanted the output to be useful beyond DATANEM itself. Excel and CSV exports let you continue with familiar spreadsheet work. SQL, JSON, JSONL and Python exports support other data workflows.

For someone preparing a report, that might mean filtering a spreadsheet by supplier or date. For a developer, it might mean loading the extracted records into a database or a script.

Being able to export matters because extraction is usually one step in a larger job. The useful outcome is a dataset you can check and carry forward into that job.

What it costs to get started

As of 27 September 2026, you can try one document without an account or payment card. A free account starts with 20 documents of credit and then provides 10 documents a month.

The Starter plan is £9 a month and includes 150 documents a month. Pay-as-you-go credit is also available at £0.20 per document, purchased in quantities of at least 25, for people who prefer to buy credit without a subscription.

These are published product allowances and prices, rather than measured performance results. The DATANEM pricing page sets out the current plans, limits and included features.

How to test it on your own work

Start with a document that represents the work you actually do. A clear invoice is a useful first test, but it is also worth checking a different layout or a typical scan before deciding how the tool fits your process.

Look at whether the important fields were found, whether the values match the original and whether the exported table is useful. For a recurring task, consider which columns you would want to keep the same each time.

I built DATANEM because I needed a better way through a pile of invoice images. My aim is to make that preparation easier for other people facing the same kind of work.

Try DATANEM with your own document or the sample invoice. No account or card is needed for the first document.

Datanem

Verified Tool

Instantly convert your documents into structured data

A
Alexander Finlay Green
Founder & Maker of Datanem · 2h ago

Discussion & Reader Feedback1

Share your teardown analysis, feedback, or discuss architecture nuances with the community.

Leave a Thought or Architecture Feedback
Posting as:
Constructive technical commentary is appreciated by makers and readers.
The SaaSearch Team
The SaaSearch TeamTeamtoday

I was genuinely impressed seeing Datanem's ability to transform those tough unstructured PDFs, scans, and images into clean, structured data for direct export into so many formats like Excel, SQL, and Python.