Live

Key-Value Extraction

Structured key-value data from any document or text.

What it does

Reads a document or its text and returns the key metadata elements as structured JSON key-value pairs — all identifiable fields automatically, or a specific list you supply. Handles PDFs (text or scanned), images, Office formats, and raw or OCR text.

Input

A PDF, image, Office document, or raw / OCR text — with an optional list of fields to target.

Output

A JSON dictionary of extracted key-value pairs.

Example use case

Auto-populate ERP or ECM fields from incoming invoices, forms, or already-OCR’d documents.

Plugs into

  • ECM metadata enrichment
  • ERP data-entry automation
  • OCR / capture pipelines
  • RPA integrations

Formats

PDF, Image, Office (DOCX/XLSX/PPTX), Text / OCR

Languages

EN, FR

Frequently asked questions

Where is Key-Value Extraction processed?+
Entirely on Nimble’s Canadian infrastructure under Canadian law — never connected to public AI APIs or foreign systems, with no CLOUD Act exposure.
Is it available in French?+
Yes. Nimble Intelligence operates in both English and French.
Can Key-Value Extraction be tailored to our documents?+
Yes. Models are custom-trained on your document types and rules, and the workflow can be adapted to your process. If you need a variation that isn’t listed, we build custom agentic workflows.

Want to try Key-Value Extraction?

Test it in our Canadian-hosted API sandbox with your own files.

Request sandbox access