How: Create a feature in the PIM system that allows users to define Inspire AI default guidelines. This feature would enable the execution of Inspire AI attributes (e.g., occasion, character, color) across a product group, but only apply specific attributes where a predefined condition or "match" is met. For example, by setting rules such as "run the occasion attribute only on collection = 'holiday'," users can control the output while avoiding manual scrubbing and filtering. Similar to a subconfig, but currently 'default' would still run on everything.
Why: The current process of running Inspire AI attributes across products is inefficient when you only need certain attributes applied to specific subsets. Without conditional defaults, you’re forced to either:
Run attributes across all styles and then manually scrub irrelevant outputs, while wasting tokens or
Run attributes individually per subset, resulting in multiple rounds of processing.
This feature will save time, tokens, reduce redundancy, and improve the quality and relevancy of automated attribute generation.
For Who: Teams leveraging Inspire AI across large sets of products
Impact: Streamlined and efficient Inspire AI processes in PIM, reducing time spent on manual scrubbing and adjustments
Enhanced relevancy and accuracy of product attributes.
Maintain consistency across products by minimizing the risk of irrelevant attribute tagging
Save thousands of wasted tokens.
Additional context:
Here’s an example of the proposed functionality:
You have 500 product styles and want to run the "Occasion" and "Character" attributes using Inspire AI.
Occasion: Should only apply to products with "Collection = Holiday."
Character: Should apply to all products, regardless of collection.
The new guideline setting in PIM would let you add default filters or rules specific to each Inspire attribute before running it across these products. Instead of processing everything and scrubbing unneeded outputs, the system would automatically skip applying "Occasion" to non-holiday products while still running "Character" everywhere.
This functionality would reduce multi-step processes and enhance team productivity while scaling Inspire AI usage effectively.
Please, add your suggestion and feedback above to add to the insights used by Product team when deciding priorities and direction of the Inriver PIM.