Products

What is Products used for?

The Products page shows the Smart Products that Nifelo built from the selected data source. You can search, filter, review and edit products, set them active or inactive, and select them for AI enrichment.

The workflow is the same for every source. This central guide is available under Data sources → General → 7. Products and every source-specific menu entry opens this same article.

This page shows the result inside Nifelo. A manual change here does not automatically change the original file, feed or store.

What is a Smart Product?

A Smart Product is the current effective product inside Nifelo. It may combine the source, Smart Field mappings, categorization, rules, normalization, supplemental data or knowledge, successful AI enrichment, and customer edits. Nifelo records the origin and ownership per field. A source-owned price can therefore update while a customer-edited title or successful AI description remains protected.

Each sellable variant may be a separate Smart Product, so the count can differ from the number of parent products in a store.

Search and filters

Filters update the list without reloading the whole page.

• Search: find a title, SKU, Smart ID or other recognizable product data.
• Per page: show 10, 20, 50 or 100 results. Actions limited to the visible page do not affect other pages.
• Category and brand: select a source product type or brand, including products where either value is missing. Use Categorization for missing Google category mappings.
• Status: Valid means active without a blocking validation error; Error means at least one blocking error; Inactive means the product is not currently active. A warning does not automatically make a product invalid.
• AI enrichment: show products for which AI eligibility is on or off. Changing this filter does not run AI.
• Price: filter by a minimum and maximum numeric price.

Click the title or price header to sort and use pagination below the table. The current selection remains represented in the URL.

Read the product table

Each row can show the main image, title, SKU, MPN, Smart ID, category mapping and source value, price, active state, AI selection and edit action. The colored left border indicates status. Hover or focus the error and warning icons beside a title to inspect findings.

The Smart ID is Nifelo's internal identity. The SKU is the commercial article code. Connected stores can also have a technical external product or variant ID. These identifiers serve different purposes and do not need to match.

Active and inactive products

An active product can participate in downstream processes. An inactive product remains stored but is not treated as active inventory. Reactivation can be blocked when the subscription product limit has been reached.

Inactive is not the same as deleted. A later complete import can affect the status when the item is present or absent in the source. Remove a product in its source when it should be removed structurally.

Select products for AI enrichment

The sparkle switch controls whether one product may be included when an appropriate AI enrichment job is started. The master switch above the column changes only the products on the currently visible page. Products on another page or outside the active filters are not changed.

The switch does not start processing and does not reserve credits. Credits become relevant only when an AI task is actually started elsewhere.

Open and edit a product

Use the edit button to open the side panel without losing the list context. The panel contains tabs for basic data, images, enrichment and raw source data.

Basic data

Depending on the source, editable Smart Fields include title, descriptions, price, sale price, stock, dimensions, product type, Google Product Category, SKU, brand, tags, MPN, GTIN, color, size, material, audience, condition and availability. A managed integration contract can make technical fields read-only. Change those fields in the source or through the intended integration workflow.

A sale price must be lower than the regular price. Nifelo also keeps related inventory and availability values consistent and blocks duplicate variant option combinations. Shared parent fields such as title, description, brand, tags, product type and category may propagate to related variants.

Category and normalization

The editor displays the resolved Google category path and ID separately from the original source value. Correct a one-off product here; correct a whole source category centrally under Categorization.

For attributes such as color, size, material, gender, age group, condition and availability, a warning may indicate that normalization is disabled or that a raw value is not mapped. Use the normalization matrix for repeating patterns.

Images

The Images tab shows the main and additional images. Fix broken URLs in the source or mapping and use Data Validation for the technical image scan. Viewing this tab does not modify images.

Enrichment and raw data

The Enrichment tab explains how the effective result was composed from imports, AI, supplemental data and rules. It does not start enrichment. Raw data shows what Nifelo received from the source and is read-only; changing Smart Data does not change that snapshot.

Saving and field ownership

On save, Nifelo compares each submitted field with the stored value. Only real changes receive customer ownership; merely opening and saving does not claim every field.

Depending on the source contract, a customer value can stay local, remain protected until that exact change is exported successfully, or be replaced by the source when it is a technical source-owned field. Relevant validation checks run again after saving.

Use manual editing for genuine exceptions. Fix repeating problems centrally: factual errors in the source, mappings under Smart Fields, group categories under Categorization, repeated values through normalization, predictable exceptions with rules, and suitable content improvements with knowledge or AI.

Recommended workflow

1. Import the current source.
2. Find priorities with Data Validation and Coverage & analysis.
3. Filter the product group you want to inspect.
4. Fix structural problems centrally or in the source.
5. Open individual products that need human judgment.
6. Review data, category, specifications and images.
7. Inspect enrichment and raw data when the origin is unclear.
8. Save only deliberate changes.
9. Select AI eligibility without expecting the switch to start a task.
10. Recheck validation and the target sales channel.