Skip to content
Nappai Documentation

Pimberly Document Loader

Pimberly Document Loader brings from Pimberly the products updated since a date and prepares, for each one, a text and a set of product details. You choose which attributes make up the text and which are kept as product details. It can also read the PDF documents linked from the product, such as data sheets or manuals.

Use it to feed a knowledge base with your catalogue, so an assistant can answer questions about your products.

When to use it

  • Build a sales assistant that knows the whole catalogue and its features.
  • Answer technical questions from each product’s PDF data sheets.
  • Update the knowledge base every week with the products that changed.
  • Prepare product texts for review or translation.

Configuration

You need a Pimberly API credential with your Pimberly account’s access key. See Credentials.

FieldWhat it does
Credential (required)The Pimberly API credential.
Since DateBrings the products updated since this date, for example 2026-01-01. If you leave it empty, it brings those updated in the last week.
Text fields separated by commaThe product attributes that make up the text, separated by commas. If you leave it empty, the text includes every attribute, each with its name.
Metadata Fields to includeThe attributes kept as product details, separated by commas. With *, the initial value, all of them are kept.
Metadata Fields to excludeAttributes you do not want to keep as details, separated by commas.
Document Fields to includeThe product attributes that hold links to PDF documents, separated by commas. If you fill it in, the component downloads the first PDF it finds in them for each product, and its content becomes the text.
Source Field TemplateHow the source of each text is built, using product attributes in braces, for example https://www.mystore.com/product/{Primary ID}. If an attribute does not exist, the source is set to Pimberly.
Split by pageWith PDF documents, creates one text per page instead of one per document. Off by default.

Advanced options

FieldWhat it does
Max ResultsThe maximum number of products brought in. Default: 10.

Output

The Data output returns one item per product, or per document or page if you use Document Fields to include. Each item carries:

  • text: the product’s or the document’s text.
  • The attributes you chose as product details.
  • source: the source, if you set Source Field Template. With documents, it is the PDF’s link, and document_type, the attribute it came from, is added too.

Connect the output to the component that stores the texts in your knowledge base.

Example

A home appliance manufacturer wants an assistant that answers with the information in its data sheets.

  1. Add Pimberly Document Loader and choose the Pimberly API credential.
  2. In Document Fields to include, enter the name of the attribute where you keep the link to the data sheet.
  3. In Metadata Fields to include, enter Primary ID,Title so each text knows which product it belongs to.
  4. Turn on Split by page if the data sheets are long.
  5. Connect Data to the component that stores the texts in the knowledge base the assistant uses.

Good practices

  • Write attribute names exactly as they appear in Pimberly and without spaces after the commas.
  • If you schedule the flow to run every week, leave Since Date empty: each run brings what changed in the last seven days. See Scheduler.