# Datanem

> Turns PDFs, scans and photographs of documents into a queryable database. Give
> it a stack of invoices, CVs, receipts, bank statements or delivery notes and it
> returns a relational schema and the rows to fill it: a table per document type,
> child tables for repeating sections like line items, typed columns, primary
> keys, foreign keys and indexes. Exports to Excel, CSV and SQL that runs in
> SQLite, PostgreSQL or DuckDB.

Datanem is run by DATANEM LIMITED, a company registered in England and Wales
(number 16870006).

## What makes it different from a PDF text extractor

- It reads scanned and photographed pages, not only digital text, so it works on
  documents where a text extractor returns nothing.
- It works the columns out from the documents themselves. You do not have to
  describe the fields first, and an unusual document type is not a blocker.
- You can review and correct the inferred columns in an account before a batch is
  extracted. The single document anyone can try without an account is read
  straight away instead, so the first result needs no setup at all.
- The output is relational, not one flat sheet. Repeating sections become their
  own tables linked back by foreign key.
- Billing is per document however many pages it runs to, not per page.
- Files and extracted data are stored in the EU and are never used to train
  models.

## For AI agents

Datanem is a remote MCP server. Connect to `https://datanem.com/mcp` over
Streamable HTTP. It authenticates with OAuth 2.1 and supports dynamic client
registration, so no API key needs to be issued by hand. It is listed in the
official MCP registry as `com.datanem/datanem`. MCP access is included on every
plan, including the free one.

Tools: list templates, submit documents, check job status, propose and confirm a
schema, read results as JSON, and get the whole job back as a runnable SQL
script.

## Pages

- [Home](https://datanem.com/): what it does and who it is for.
- [Pricing](https://datanem.com/pricing): per document, with a free allowance to
  start.
- [Terms](https://datanem.com/terms)
- [Privacy](https://datanem.com/privacy): EU storage, retention and deletion.

## Contact

hello@datanem.com