A new supplier sends a thousand products with a code, a price and a few photos, and someone has to write the descriptions, categories and translations by hand into PrestaShop or WooCommerce. AI prepares the whole batch following your style rules and saves it as drafts, and a merchandising manager reviews and publishes.
Built for ISO 27001NIS2 DORA
A typical situation in an online shop: a supplier sends an Excel price list with product codes, names in the manufacturer's language, a few attributes and a link to a photo folder. In the PrestaShop or WooCommerce admin, a merchandising manager comes up with a title for every product, writes a description, assigns a category, puts the attributes into the right fields and translates everything into the shop's languages. It takes 10 to 20 minutes per product, so a batch of a thousand products turns into a month of work or into half-empty product cards. The same products also need to go to the Amazon marketplace with its category tree and to Google Merchant Center for distribution, where empty attributes mean rejected products.
The solution works in batches. We connect Claude through the API to your PrestaShop, WooCommerce or Shopify product API: the agent takes the supplier price list, the attribute table and the photos, prepares a title and description following your style guidelines (tone, length, mandatory fields, forbidden claims), determines the category from your tree using the attribute table and the photo, fills in the attribute fields and then translates into the chosen languages. From the same data it prepares an Amazon import file using the marketplace categories and completes the Google Merchant Center feed fields. Each product gets a confidence flag: where data is missing or the attributes contradict the photo, the agent marks the item for review. The manager reviews the batch in a review window, corrects what is needed and approves, and the products are written to the online shop through the integration as drafts or published immediately. The model runs on the Anthropic API or Amazon Bedrock in an EU region, your data is not used to train the model, and every batch and its changes stay in the log.
Four steps from input to a record in your system, with human approval wherever there is risk.
The manager uploads the supplier's Excel price list, the attribute table and the photo folder, or points to the supplier's data feed. The system maps the columns to the fields of your PrestaShop or WooCommerce product card and shows what is missing before any writing starts.
For each product Claude prepares a title and description, assigns a category from your online shop's tree and fills in the attributes from the table and the photo. It then translates into the chosen languages, keeping the terms from your glossary, and prepares the Amazon and Google Merchant Center fields.
The batch appears in a review window with doubtful products flagged. The manager corrects what is needed and approves, and the products are written to PrestaShop, WooCommerce or Shopify through the API as drafts or published immediately, depending on your setting. The Amazon file is exported separately.
If the manager keeps correcting the same thing, we update the style guidelines or the category map so that the next batch arrives cleaner. Products with missing data stay on a list to request from the supplier, and products rejected by Merchant Center are returned to the batch.
Writing products, categories and attributes to PrestaShop, WooCommerce or Shopify through the product API, in draft or publish mode, and an import file for Amazon.
Rules prepared together with you: tone, length, mandatory fields, forbidden claims and a term glossary for every language, aligned with Google Merchant Center requirements.
The place where the manager sees the whole batch and the flagged doubts, and can edit and approve products one by one or all at once before they are written to the online shop.
A record of every batch showing what was generated, what was corrected and what was published, and our support when a PrestaShop or WooCommerce version is updated or supplier formats change.
Claude via API + PrestaShop, WooCommerce or Shopify product integration, optionally an Amazon import and Google Merchant Center fields. We connect through the platform API, a file export or an email flow, depending on your version. Read more about this service and its limits on the service page.
The agent writes only from the attributes the supplier provided and what is visible in the photo, and our style guidelines forbid claims without a source. When data is missing, the product is flagged for review instead of being filled in by guesswork. The manager always approves the final text before publishing in PrestaShop or WooCommerce.
Yes, all three have a product API through which we write products, categories and attributes. The differences are in versions and plugins, so we choose the integration method (API, CSV import or email flow) after checking your platform in the first conversation. Custom-built platforms we connect through their import files.
No. We work through the Anthropic API or Amazon Bedrock in an EU region, where your data is not used to train models. If needed, descriptions can be generated with a private model in our GPU cloud in Lithuania (EU).
Reviews land on the website, Google, Amazon and eBay, while return reasons sit in a free-text field of the SAP Business One return document that…
A new cashier at the Lightspeed till does not know whether a return without a receipt is allowed, what promotion runs today or what to do with a…
A supplier invoice, a delivery note, the SAP Business One order and the warehouse receipt rarely match, and the differences surface only after…
Customers ask whether an item is in the store, when an order will arrive and how to return it, while the support team checks the same things in…
This service is delivered by specialists with valid international certifications. Their expertise is confirmed by independent organisations, not by us.
AI systems are deployed with protection against data leakage and model abuse.
The AI security programme and its controls are managed following the ISACA methodology.
AI risks are identified and managed like any other IT risk.
We reply within 1 business day
