AI product content for eCommerce
AI Data Generator
Turn raw product data from an Excel file into on-brand, eCommerce-ready product content. Import, generate, export.
AI Data Generator
Raw row from your Excel file
- sku
- TC-2041
- title
- trench coat dbl brst M
- material
- CTN100%
- colour
- SND-01
-
Normalization
- material
-
CTN100%becomes 100% Cotton - colour
-
SND-01becomes Sand
-
Segmentation
- Product type Trench coat
- Colour Sand
- Material 100% Cotton
- Size M
-
Generation
- en_name
- Double-breasted cotton trench coat in sand uses title, material, colour
- en_short_description
- A belted, double-breasted trench coat in 100% cotton, cut to knee length. uses en_name, photo
- en_meta_title
- Cotton Trench Coat in Sand, Double-Breasted and Belted uses en_name, en_short_description
-
Translation
- rs_name
- Dvoredni pamučni kaput boje peska uses en_name
- rs_meta_title
- Pamučni kaput boje peska, dvoredni, sa kaišem uses en_meta_title
Trusted by retail teams
Why it matters
From thin supplier data to content that sells.
Product data rarely arrives ready for a shop. The AI Data Generator turns it into content your shoppers, your filters and search engines can use.
Without it
- Supplier files arrive with thin titles, codes and abbreviations such as “CTN100%”.
- The same colour or material is written in several ways across the catalogue.
- Names, descriptions and meta fields are written by hand, product by product, in every language.
- Attributes are missing or inconsistent, so filters and search miss products.
With the AI Data Generator
- Raw values become consistent terms: “CTN100%” becomes “100% Cotton”, everywhere.
- Names, descriptions and meta titles are generated in your tone of voice, following your rules for categories and brands.
- Content is translated for each market, so every language version is complete.
- Clean, structured attributes such as colour, material and size improve filters, search and SEO.
How it works
From an Excel file to finished content, in four steps.
You set the rules once; every file you import follows them. The demo above shows what happens to one product row.
Configure fields and groups
Choose the fields to generate, how and in which language, and bundle them into a Data Group per market, language or category.
Import a file and generate
Import an Excel file with your products, choose the group and click Generate. The fields run in priority order.
Review and correct
Check the generated values and correct anything that needs it. Every edit is logged.
Export
Export the finished file: your original columns plus the generated fields, ready for your shop or PIM.
In the dashboard
How your team does it in Athena
Five steps in the AI Data Generator section of the Athena dashboard, with the labels you will see there.
1. Configurator
Tell the AI what your file contains
- In the sidebar, open AI Data Generator → Configurator.
- In Data Mapper, click Add Field for each column of your file: its name in the file, its name in the system and its type.
- Mark a column Is Excluded when its value should stay in the file but not be used as context for the AI.
- In Data Groups, click Create Generator Data Group, for example one group per market or language.
Recreated dashboard screen
Configurator
Instruct the AI so it will work the way you need it to.
- Data Configurator
- Data Groups
- Data Mapper
- Image Prompt
- System Prompt
- Knowledge Base
Columns
- Output Name
- Input Name
- Field Type
- Field Description
- Is Excluded
- Actions
Field Type
- Numerical
- Text
- Url
2. Fields
Create the fields you want generated
- In Data Configurator, click Create AI-Generated Field.
- Pick the Agent Type: Normalization, Segmentation, Generation or Translation, then write the field’s Prompt: structure, tone, length and terminology.
- Under Field Awareness, choose the columns and fields the AI should see, and set Priority so the field runs after the fields it uses.
- Add the field to its Groups and save.
Recreated dashboard screen
Configurator
Create AI-Generated Field
- Data Configurator
- Data Groups
- Data Mapper
- Image Prompt
- System Prompt
- Knowledge Base
Agent Type Generation
Name en_name
Status Active
Use Image Output No
Priority 0
Highlighted: Field Awareness title, material, colour
Prompt Product name in English: product type, material and colour.
Creativity Set per field
Effort Set per field
Groups English market
Prefix / Suffix Optional
3. Import and generate
Import a file and click Generate
- On the AI Data Generator main page, click Import the file.
- Click Choose File to pick the Excel file with your products, then Choose Group… to pick the Data Group for it.
- Click Generate. The active fields of the group run over the file in priority order.
- Picked the wrong group? Stop Import ends the run.
Tip: try a small sample file first, and include as many descriptive columns as you have. The more context the AI gets, the better the content.
Recreated dashboard screen
AI Data Generator
Generate product content with AI.
Choose File products.xlsx
Choose Group… English market
4. Review
Review, correct and see every change
- Open Generated Data and select a processed file to preview its results.
- Correct any value that needs it.
- AI Data Generator Edit Logs records each change: the original and updated value, the difference, the field, the file, who changed it and when.
Tip: when the same field keeps being corrected, change its prompt or add a Knowledge Base entry instead of editing row by row.
Recreated dashboard screen
AI Data Generator Edit Logs
Columns
- ID
- Original Value
- Updated Value
- Highlighted: Differences
- Updated Field Name
- File ID
- Admin
- Product Row ID
- Updated At
Updated Field Name en_name
Original Value Cotton trench coat in beige
Updated Value Cotton trench coat in sand
Differences beige → sand
5. Export and Reports
Export the file and follow the usage
- On the main page, click Export to download the file: your original columns plus the generated fields.
- Open AI Data Generator Reports and pick a date range to see how many products, SKUs, images and words were processed.
Recreated dashboard screen
AI Data Generator Reports
Date filter
- Today
- Yesterday
- Highlighted: Last 7 Days
- This Month
- Last Month
Metrics
- Products Processed
- Unique SKUs Processed
- Regenerate Clicks
- AI API Calls
- Images Processed
- Average Number of Words per Data Type Field
- Number of Input Words
- Number of Output Words
- Data Type Fields Generated
- Average Number of Characters per Data Type Field
- Number of Input Characters
- Number of Output Characters
Products Processed over time
- 1h
- 3h
- 6h
- 1d
- 1M
Simplified recreations of the dashboard screens; labels as in the dashboard. Everything is scoped to the store you select at the top of the dashboard.
Capabilities
Everything in the AI Data Generator
Every product data capability on the Athena feature list, and who uses it.
AI Data Generator
Turns an Excel file of raw product data into eCommerce-ready content: import, generate, export.
Who uses it: Catalogue and content team
On-brand content
Tone of voice and writing style are set per store, so the content reads like your brand, not like generic AI text.
Who uses it: Catalogue and content team
Content generation
Names, short and full descriptions, meta titles and meta descriptions, and custom fields you define.
Who uses it: Catalogue team
Translation
Translates product content into the languages of your markets.
Who uses it: Catalogue team
Segmentation and classification
Classifies products by your predefined rules and reference attributes, such as colour, material, size and product type, for clean, filterable attributes.
Who uses it: Catalogue team
Normalization
Standardises inconsistent values, so “CTN100%” becomes “100% Cotton” and your catalogue uses the same terms everywhere.
Who uses it: Catalogue team
Fields and Data Groups
Define the fields to generate and group them per market, language or category, then apply a group to each file.
Who uses it: Catalogue team
Data Mapper
Maps the columns of your file to system fields, so the relevant data is recognised.
Who uses it: Catalogue team
Prompts per field and per agent
A prompt for each field and a System Prompt for each agent type control terminology, tone and outcome.
Who uses it: Catalogue team
Computer Vision
The AI reads what a product image shows and turns it into image-related text. Images are used only as input for the AI.
Who uses it: Catalogue team
Field Awareness
Any generated field can be the input of another, so fields build on each other without repeating themselves.
Who uses it: Catalogue team
Generated Data review
Preview, review and correct the results of each file before you export them.
Who uses it: Catalogue team
Edit Logs
A full audit trail of generated values and of every manual correction.
Who uses it: Catalogue team, compliance
Generator Reports
Processing metrics, such as products and SKUs processed, AI calls, images, and word and character volumes, with date filters.
Who uses it: Catalogue team
API access
The AI Data Generator is also available through the API, so content can be generated without manual imports and exports.
Who uses it: Catalogue team
AI Data Enricher
Enriches an already-published catalogue with the same AI pipeline and finds missing data.
Who uses it: Athena team
The Athena team runs it for you.
AI Data Enricher
For the catalogue you already have online
The Athena team runs it for you.
The AI Data Generator works on the files you import. The AI Data Enricher works on products that are already published, with the same configurable AI pipeline.
- The same fields, Data Groups, prompts and Knowledge Base as the AI Data Generator.
- Finds the data your published products are missing.
- Search can use enriched fields automatically.
- Enriched fields can drive automatic product boosting in search.
From catalogue to search
- Your published catalogue Products already live in your store.
- The same AI pipeline Fields, Data Groups, prompts and Knowledge Base.
- Missing data found The Enricher finds the data your products are missing.
- Used by search Search can use the enriched fields, including for automatic product boosting.
FAQ
Questions about the AI Data Generator
Do we import and export Excel files?
Yes. In the dashboard you import an Excel file (.xlsx) with your raw product data, choose a Data Group and click Generate. Export gives you the finished file: your original columns plus the generated fields. The AI Data Generator is also available through the API.
Can it write content in several languages?
Yes. Create a Translation field for each language, or a Generation field written directly in the target language. A good order is to generate your main language first and then translate the generated fields.
Can one field use another field’s output?
Yes, with Field Awareness. Add the source field to the dependent field’s Field Awareness and set Priority so the source runs first. That is how a Serbian meta title can be translated from the English one generated before it.
Can it use our product images?
Yes, as input for the AI. Set Use Image Output to Yes on a field and configure an Image Prompt, and the AI turns what the photo shows into text, such as a belted, double-breasted cut. The AI Data Generator does not edit or process the image files themselves.
How do we keep our terminology and tone of voice?
The content follows your instructions: your tone of voice and your rules across categories and brands, which you provide. The Knowledge Base maps terms, such as a supplier’s word to yours, and a System Prompt per agent type sets the standing rules for every field of that type.
What is the difference between the AI Data Generator and the AI Data Enricher?
The AI Data Generator processes the files you import. The AI Data Enricher works on the catalogue that is already published in your store, with the same pipeline, and the Athena team runs it for you.
See what Athena can do for your business.
Explore your catalogue, integration needs, and opportunities to improve product discovery with our team.
Developed by Syncit Group, certified to ISO 9001 and ISO/IEC 27001
GDPR-compliant
Hosted on AWS in the EU, running 24/7
Typically live in 2–3 weeks