How IDP is organised
Intelligent Document Processing (IDP) reads documents, images, text and voice notes, works out what they are and turns them into structured data.
You configure it in layers, from general to specific:
| Level | What it defines |
|---|---|
| General recognition options | Rules that apply to every document the assistant handles, plus the base language and an optional final workflow. |
| Document types | Categories such as “Supplier invoice” or “Daily report”, with their own rules and image corrections. |
| Page layouts | A specific visual template, based on a sample image, used to recognise pages and target areas. |
| Areas | Parts of a layout with precise instructions for a field or table. |
Step 1: Set general recognition options
In the assistant’s context menu, open General Recognition Options.
In AI Instructions, write rules that apply to all documents – which data to extract, how to treat dates and amounts, and what not to invent if it’s missing.
Set a Base Language for IDP if documents arrive in several languages and you want consistent output.
Optionally choose a Final Workflow to run after recognition – for example to send the data to your CRM or email a summary.
Step 2: Create document types
Open Document Types and click Add.
Give it a specific description, such as “Materials delivery note”.
Add Custom AI Instructions: where to find fields, how to read tables and rows, and what to ignore.
Tick Use General Rules to apply the assistant’s general rules too.
Only enable automatic corrections – Redraw Grid, Fill Empty Cells, Auto Rotation – if they genuinely help.
Step 3: Create page layouts
Open Page Layouts for IDP and click Add.
Link the layout to its Document Type and choose the Page Type – for example first page, middle page or last page.
Upload a clean, complete sample image of the document.
Draw areas over the fields you need, give each a Field Name and, if useful, specific AI instructions or a rotation.
Tick Use Document Type Rules if the document type’s rules should apply as well.
Good to know
- Create separate layouts when the structure changes – for example a cover page, detail pages and a summary page.
- Use Discard This Page for pages such as terms and conditions that should be recognised but not extracted.
- To carry data such as a document number or site from the first page to later pages, describe it in the crosspage_data field in your rules.
- Only enable Use Defined Areas Only when fields always appear in the same place.
Step 4: Add validation fields
Validation fields correct and encode values – for example mapping “M. Rossi” or a misread “Marco Rosso” to the resource “Marco Rossi” with code RES001.
Open IDP Validation Fields and click Add.
Choose the Field to validate, such as resource, job site, unit of measure or payment method.
In Parameters, add one CSV row per value: the code, the standard description and then any aliases – for example “RES001”, “Marco Rossi”, “M. Rossi”, “Rossi Marco”.
Next, test your setup with the IDP processing test.
Prefer a guided walkthrough? Book a training session and an Echo7 specialist will help your team set it up.
Book a training sessionWant to go deeper?
This guide covers the essentials. For detailed reference on every setting and feature, explore the full Echo7 Wiki.
Echo7 Wiki