Overview

Already have a data model in your PIM and need to onboard a new product line? The ‘Catalog Expansion’ mode lets the Data Architect Agent build on top of your existing catalog structure, keeping your current families, attributes, reference entities, and asset families intact while suggesting new ones tailored to your uploaded data.

 What Catalog Expansion does:

• Generate entities for new product lines (families, attributes, reference entities, asset families).
• Reuse and extend your current structure (e.g., apply existing modeling patterns/attributes when creating new families).
• Keep your current model safe: existing entities aren’t modified or deleted.
  
   

When should you use the ‘Create New Model’ or the ‘Expand Existing Model’:

Start by creating your data model
 

How does it work?

• Check that you have full permission access to the Data Architect Agent under System > Roles > Administrator (or your own role) > Data Architect Agent
   

• Then, click onto Settings to find the Data Architect Agent under the Automation section, all the way to the bottom of the page and select ‘Expand PIM structure’.

• Name your data model and select your locales

Start by giving your data model a name, then define the locales the Data Architect Agent will use to generate your catalog structure. This configuration ensures that all generated families, attributes, and options are properly labeled in the right languages, ready for your team to use and consistent with your PIM setup.

You have the possibility to choose several locales to generate entities labels, but only one for the technical codes. 

• Locales to generate entities labels

This setting determines the labels for your generated entities.You can select multiple locales (e.g., English, French, German etc.) to generate labels in several languages simultaneously. These labels will appear in the corresponding languages for your families, attributes, attribute options, reference entities and assets.

• Locale used for technical codes

This setting identifies the primary language used to generate the unique technical codes for your entities. Codes are typically generated in a single language (often English) to maintain a clean and predictable naming convention for your database and API integrations.

• Fill in the ‘Advanced Prompt’ text box to guide the AI model and get the most of it. 
  
  The Advanced Prompt field allows you to provide additional context to guide the AI when generating your data model. While the Data Architect Agent can build a structure without it, adding a prompt helps the AI better understand your business specificities and produce more accurate, tailored results.
   

Upload your files
 

Then, upload the product data file(s) that have been extracted from your ERP or your e-commerce platform. Below is an example of the kind of data structure we expect to receive:

Family Composition

With the new Catalog Expansion, we’ve introduced several family modeling strategies to ensure the DAA aligns with your specific business logic. 
When applying your model, you can now choose between three levels of granularity to ensure these new product lines will be structured exactly how you want: 

• Consolidated (fewer, broader families)
• Balanced
• or Granular (highly specific, detailed families)
   

Key takeaways from Catalog Expansion

The DAA follows a strict "non-destructive" policy: it will not modify or delete your existing attributes or families. When you click ‘Apply’, the agent only creates the new entities proposed in your current DAA model. 

• Conflict resolution: 
  • If the DAA suggests an attribute code that already exists in your PIM, the PIM version will always be the source of truth and there will be no automatic overwrite. In the event of a conflict between your live PIM structure and the current DAA modeling, conflicts will be identified under the ‘Modeling’ tab for your resolution. You will be able to ‘Resolve’ these issues by choosing among three options :
    • Rename the attribute label in the DAA model
    • Remove the attribute from the DAA model
    • Use the attribute from the existing PIM structure
       
• Reference entities & Assets: Existing asset families and reference entities are preserved. The DAA only generates new ones if it’s necessary for the new product line.

Handling families & validation

• Existing families: If the DAA suggests adding new attributes to a family you already have, those attributes will be added to that family as soon as the model is applied.
• Validation rules: The Data Architect Agent only runs validation checks on the newly suggested entities. We assume your existing PIM data is already valid, so the DAA won't flag errors on your pre-existing attributes.

Explore and refine

After processing, the tool creates a data model with families, attributes, reference entities and options. You can explore and make adjustments to every entity generated (families, attributes, reference entities, codes and labels etc.). 

The data model generated by the DAA is a starting point, not a final output. We strongly recommend refining it with your business and catalog expertise to ensure it accurately reflects your needs before proceeding.

Reviewing suggestions

To review existing entities, you can easily check them under the ‘Family Assignment’ tool, accessible from the ‘Configuration’ tab. You then have two different choices, to either show the new attributes from the DAA model or to show a consolidated view of both new attributes from the DAA model and the existing PIM structure.
 

Source File

The 'Source File' tab to provide greater visibility between the input files' columns and the data model generated by DAA, to help you understand DAA's modeling decisions.

Audit

The ‘Audit’ tab is designed to give you a comprehensive overview of the data model generated in the DAA. This feature breaks down key metrics, including attribute distribution and properties, and gives you a detailed view of both shared and family-specific attributes.

NEW - Collaboration

The DAA supports collaborative work on a data model: you can now edit and review data models created by other users in the same instance.
 

To help you keep track of changes, any modifications made by other users are clearly highlighted in the History tab of your data model with a visual indicator, so you can quickly identify what has been updated without having to review the entire model.

A visual indicator keeps you informed of how many changes other users have made to the data model. You can review these changes at any time and restore a previous version if needed.

Key capabilities and support

Product previews

The Product previews tab gives you a first overview at your products, showing you how they'll appear in the PIM.

We generate a sample set of products from your uploaded file, based on the families and attributes that have been suggested. You can easily navigate through your products by using the family filter or the search bar.

The attribute preview is designed to give you a quick projection of your data, by creating a sample set of your products.

Please note that:

• The preview currently supports text, text area, simple select, and multi select attributes. Other attribute types may not be displayed as expected in the preview.
• The product preview does not guarantee that every attribute will have a value, nor that every line in your uploaded file will necessarily result in a product in the preview tab.
• The product preview might take some seconds to load.

Importing the data model

When you are satisfied with your data structure, you can import your model into your PIM to create your entities, using the ‘Import your data model’ button on the top right of the screen. 

You have three choices when importing the data model:

• Importing the model only: This will import the families, attributes, reference entities, and assets, without any products.
   
• Importing the model and a product sample: This will import the families, attributes, reference entities, assets, and a sample of your products (approx. 20 products for the moment). This provides users with an immediate, visualized first version of their catalog structure, accelerating feedback and project iteration. 
   
• Importing the model and create a Tailored import profile : This will import the families, attributes, reference entities, assets. It also creates a Tailored import profile, which gives you complete visibility into how your source file columns map to the generated PIM attributes. You can then refine and save this mapping for maximum accuracy and future product imports.
  
  
  On this screen, confirm your choice to initiate the data model import and profile creation. 

If your select the third option, you will have a modale screen where you will be able to access the 'Tailored Imports' tab in the Data Architect Agent when the processes finish. You have two actions to choose from ‘View Import Structure’ or ‘Launch Import’.

• If you select ‘View Import Structure’, you will be redirected to your Tailored Import Profile to configure your Product Mapping:

On this page, you can review one last time how the mapping has been done by checking the columns of your input file (the Source) and the attributes in your PIM (the Target). You can also review and/or add operations to improve the data quality and minimize errors during import. 

• Conditional settings : depending on the target you selected, some other parameters need to be selected:
  • For example, for scopable and/or localizable attributes, you must select to which channel and/or locale you want to import your data.
    
     
• Advanced mapping operations: 
  You can use operations to transform the data in order to make the import process easier.

• Asset Mapping:

If your uploaded file contained media links, you can use our powerful new Asset Mapping to convert those entities into PIM 'Assets'. This new addition is divided into two panels. 

• Asset Mapping Structure (Left panel): 
  • In the left panel, the ‘Asset mapping structure’ selects the Asset Attribute (the target) that will link the product to the corresponding assets.
  • You can select the same Asset Attribute multiple times, enabling a single product attribute to link to various assets.
    
    

    Scopable and localized attributes are fully supported in the Asset Mapping structure.

     

• Asset Configuration (Right panel):
  • Asset code configuration: Define the strategy for how the final asset code will be generated. You can choose between two strategies.
    • Product identifier + suffix : The asset code will use the product’s main identifier, or the import ID if the main identifier is empty. A suffix is optional.
    • Column value : The asset code will be the value of the column selected. 
  • Attribute mappings: Map the asset family's codes to specific columns in your source file. If you already have a column with the asset code in your uploaded file(s), this will ensure that assets are created with the necessary information.
     

    If you need to map several URLs from your input files in a single asset family, you can create different mappings for the same asset family. 

     

• Reference Entities Records Mapping

You can now define how new Reference Entities records will be created and linked to your products during the import process. 

First, select the relevant Reference Entity attribute, then specify how the source file data should populate each record field.

• Record Mapping Structure (Left panel): 
  • In the left panel, the ‘Record mapping structure’ selects the Reference Entity Attribute (the target) that will link the product to the corresponding records.
  • You can select the same Reference Entity Attribute multiple times, enabling a single product attribute to link to various records.
    
    

    Scopable and localized attributes are fully supported in the Reference Entities Records Mapping structure.

     

• Record Configuration (Right panel):
  • Record code configuration: Define the strategy for how the final record code will be generated. You can choose between two strategies.
    • Product identifier + suffix : The record code will use the product’s main identifier, or the import ID if the main identifier is empty. A suffix is optional.
    • Column value : The record code will be the value of the column selected. 
  • Attribute mappings: Map the record attribute's codes to specific columns in your source file. If you already have a column with the record code in your uploaded file(s), this will ensure that records are created with the necessary information.

• Next action to import your products: launching the ‘Quick Import’

Before proceeding, we strongly recommend reviewing the structure of the generated mapping. For each PIM attribute (Target), you can select one or multiple sources from your input file, depending on the attribute type. 

At this stage, your data model structure (Families, Attributes, Reference Entities and Assets) has been already created in your PIM but you will need to select the ‘Quick Import’ button (top right of your screen) to import your products. 

You then have two choices, either:

• Using the current file provided during the DAA modeling process
• Or, you can also upload a new file, which will be imported using the saved structure. This option supports both testing with smaller datasets to validate the structure, and scaling up to import larger datasets (bigger than the DAA's original file for ex.).

Additional capabilities

Data Architect Chat

Get instant guidance on your Akeneo PIM data modeling. Use the chat to ask questions about best practices and receive advice on your attribute definitions.

Please note: This experimental tool provides guidance only and does not have the permissions to modify your live taxonomy or data model.

Access & Permissions : To ensure a consistent experience, only the original creator of the data model can interact with the chat. All other users can view the conversation in read-only mode.

Conversation Context : Each chat is unique to its specific data model, allowing you to keep your modeling discussions organized. While the full context of your conversation is preserved indefinitely, the chat interface will display the last 100 messages in your history for optimal performance.

Download

In the top right corner of the DAA, you'll see additional actions, including the ability to download your files. 

You have several options to choose from, depending on your needs.

•  The full Data Model in a XLSX format: this option provides a single Excel file containing a comprehensive overview of the data model that has been generated in the Data Architect Agent.
  
  The file contains dedicated sheets for Attributes, Families, Family variants, Attribute groups, Asset families, Reference entities, and Select options.
  
  We have also included sheets for Attribute definitions, Metric Family codes, and Metric units. These are intended as technical guides for your modeling and are not part of the importable structure.

• If you would like to re-export these files, you have the possibility to download individual files in CSV formats for the following entities:
  • Attributes
  • Attributes Options
  • Families
  • Family Variants
     

Specifications

• Depending on several factors, generating a data model can take anywhere from minutes to hours. A waiting message “Model generation in progress” will be displayed on your screen, and the page will be automatically refreshed to show the model generated. A confirmation email is also sent when the model has been generated.
• The LLM used for data model generation is Gemini Flash 2.5. Data sent to Google is neither stored on their servers nor used to train their models.
   

What kind of context should I give to the AI model?

It is important to note that custom instructions always take the priority over a default behavior. It means that if you write something specific, even if it breaks usual PIM rules, the model will follow it. 

The goal is not necessarily to be fully exhaustive, but it is to remove uncertainty and make the output fit better with your customer’s business, tools and their way of modeling data.

If you don't give any instructions in the ‘Business & Data Modeling Context’, the AI will still work, as:
- It understands how the PIM works
- It applies best practices by default
- It reads and understands the structure of your input product file
- It produces a valid, coherent taxonomy using standard logic
 

Keep in mind that prompts need to be simple and easy to understand.

Examples of Custom Instructions you can use (with rationale)

There’s no fixed structured that we would particularly recommend but you can find example prompts you can use to make the model behave exactly how you want - to improve consistency, resolve ambiguities, and align the results with the logic of your catalog, your tools or your business.

Use the variant_axis column to define variant products.

Rationale:

The model can infer good variant axes from your product data — for example, if it sees multiple products with identical names but different colors, it may suggest color as a variant axis. But if you want to enforce a specific structure or match another system (like an ERP or spreadsheet you’re importing from), you can specify it directly. This is especially helpful when your file contains edge cases that could mislead the model, like inconsistent naming or overlapping dimensions.

Treat the details field as a select-type attribute.

Rationale:

If the details field contains recurring values these could be modeled as options in a simple select attribute. But if the field has a lot of different values and not many of them are recurring for different products, the model might otherwise treat it as free text. Adding this instruction helps resolve ambiguity and structure your data better.

I want the data model to include the following families: spare_parts, retail_accessories, e_bike_components.

Rationale:

The model usually creates families by analyzing product clusters, but some families might not be created as you want them, or if the model chooses a more granular or more general group instead. Listing the families here ensures they appear as-is. This is also helpful if some products that belong to these families are missing from your file (e.g. partial exports or subset samples). It also helps resolve granularity mismatches (e.g. if you want “retail_accessories” to be a consistent macro-family, even if the model would have split it into “bottle_holders”, “lights”, etc.).

I plan to add a new outdoor_gear product line to my catalog in the future, so create a family for it that should mirror retail_accessories.

Rationale:

This helps to future-proof the structure. The model won’t generate structure for products that aren’t in your file, but if you know you’ll add a product line (like “outdoor_gear”) you can mention it. You can also give the model a reference: saying it should “mirror retail_accessories” tells the model to use the same kind of attribute set, variant logic, etc. This is useful when retail_accessories is well structured in your file and you want the new family to inherit its logic.

Use ALL CAPS for attribute labels. Limit them to 120 characters.

Rationale:

This enforces internal style conventions. While best practices would usually lean towards sentence casing (i.e. capitalizing only the first letter of the first word in a sentence and any other proper nouns), you have the possibility to override to fit other conventions. The model will mostly follow the constraint (especially for casing), but character limits may not always be perfectly respected - still, setting the expectation ensures labels are short and clean in most cases.

Attribute codes should include the unit at the end (e.g. weight_g, length_mm) and the label should match: “Weight (g)”.

Rationale:

In PIM best practices, including units in codes or labels is discouraged as units are usually handled using the ‘Metrics’ attribute type. But if your systems (e.g. Excel exports, ERP schemas) require it for clarity or consistency, you can enforce it using custom instructions. The AI will follow this logic even if it wouldn’t propose it by default.

There should be an “Unknown” option for the material attribute.

Rationale:

The model usually expect you to leave attributes empty when products are missing a value for this attribute. If your system requires placeholder values for completeness, or if blank fields break exports or validation rules, you can enforce fallbacks.

Example of a full Custom Instruction

I’m a B2B distributor. Use the product_type column to define families and variant_axis for variant grouping. Attribute codes should always end in a unit (_g, mm), and the labels should reflect that (e.g. “Weight (g)”). Group all technical attributes under a tech prefix. Do not introduce any new attributes that aren't already in the file. For when material is missing, add an “Unknown” option.

Discover Attribute Suggestions from Activation & PX Insights

Change log of our latest improvements