Docs  /  Building a Configurator  /  AI Agent: Import Product Data

AI Agent: Import Product Data

Last updated: 01. January 2026

The AI agent can read product data from various sources and import it directly into your configurator — CSV files, Excel spreadsheets or structured text lists.

Opening the AI Agent

Open a component in the V3 backend. Click + New Chat at the bottom right to open the chat.

Supported Formats

CSV and Excel

Upload a CSV or Excel file directly in the chat. The agent automatically recognises the structure and asks for clarification if needed.

Example — option list as CSV:

Feature;Option;Surcharge
Colour;White;0
Colour;Anthracite;0
Colour;RAL custom colour;45
Material;Aluminium;0
Material;Stainless steel;80
Material;Wood;35

"Here is my option list as CSV. Please create the features and options accordingly."

Price Lists as a Table

;80cm;100cm;120cm;140cm
100cm;29.90;34.90;39.90;44.90
120cm;34.90;39.90;44.90;49.90
140cm;39.90;44.90;49.90;54.90

"This is my price list. Columns = width, rows = height."

Structured Text

"My product has the following features:

  • Width: free input, 50–400 cm, required
  • Colour: selection — White (€0), Anthracite (€0), Bronze (+€25), RAL special lacquer (+€60)
  • Drive: selection — Manual (€0), Electric 230V (+€95), Solar (+€140)
  • Base price: €49"

Importing Existing Data from Other Systems

From a PIM System

Export your product data from the PIM as CSV and upload the file. Tell the agent which columns correspond to which features.

From an Excel Price Calculation

Upload the spreadsheet — the agent recognises common structures and imports the prices as a price matrix (under the Price Lists section in the structure tree) or surcharge list.

Updating Data

"Here is my new price list for 2026. Please replace the existing price matrix."

"The option 'Anthracite' should now be called 'Anthracite grey RAL 7016' and have a €5 surcharge."

"Add the following new options to the colour list: Silver grey (+€0), Dark brown (+€0), Cream white (+€0)."

Tips

Clean source data: No empty rows, consistent separators, clear column headers.

Check encoding: CSV files should be UTF-8 encoded so that special characters are imported correctly.

Import step by step: For complex products: first features, then prices, then rules.

Next Steps


← All articles Try free for 14 days