AI image

How to Onboard Supplier Product Data into Pimcore (2026)

Pimcore's Data Importer, Pimcore Store extensions or an AI onboarding layer: how to get supplier PDFs and spreadsheets into your Pimcore classes, checked.

Zofia Kamińska AI character Reviewed by Olgierd Mrozik
9 min read Updated 13 October 2026

Disclosure. openProd publishes this page, and openProd is one of the tools described. Pimcore and every other vendor are described only with what they state in their own documentation or listings, with a link to the source. Checked on 13 October 2026. If something has changed, tell us and we will update this page.

The short answer. There are three routes for supplier product data into Pimcore. Pimcore’s own Data Importer (part of Datahub) is enough when suppliers send clean CSV, XLSX, XML or JSON files that match your classes. Extensions from the Pimcore Store, such as Onedot Product Data Onboarding, add AI classification and catalogue import. An onboarding layer in front of Pimcore, such as openProd, reads PDFs and spreadsheets and maps values to your model before anything reaches Pimcore.

Three routes into Pimcore

  1. Pimcore’s own tools. The Data Importer reads structured files and writes them into data objects through a mapping you configure. Pimcore workflows then move products through your own review steps.
  2. Extensions from the Pimcore Store. Third-party products listed in the official Pimcore Store that add AI to onboarding, classification or enrichment.
  3. An onboarding layer in front of Pimcore. A separate tool where your team brings in supplier files, checks what the AI read, and then hands the result to Pimcore.

The routes are not exclusive. Many teams keep the Data Importer for suppliers who already send clean files and add a second route for the rest. The question is which files need a person today and how much of that work you want to move.

What the Pimcore Data Importer does

The Pimcore Data Importer adds import to Pimcore Datahub. According to its documentation it reads from six data sources (asset, upload, HTTP, SFTP, push endpoint and SQL) in five formats: CSV, JSON, XML, XLSX and SQL query results. You build the mapping in an editor with a live preview, so you see source data and the result of each transformation step before you run it.

Two more features matter for supplier files. A delta check skips records that have not changed since the last run, and a cleanup option can remove data objects that no longer appear in the source. Imports run manually, on a schedule, from the command line or when a push request triggers them.

The transformation pipeline turns source values into the target format with operators: trim, string replace, combine, explode, numeric, date, quantity value, asset import and more. A Conditional Conversion operator maps string values onto other string values, which helps when a supplier’s wording differs from yours.

PDF is not a data source or file format in the Data Importer documentation, and the documentation does not mention AI. The Data Importer is built for structured data: it imports reliably once someone has defined how the columns map.

What a supplier file needs before an importer can use it

A supplier data sheet is written for a buyer, not for your Pimcore classes. Before a file can go into any importer, someone has to do four things.

  • Read the data. Product names, specification tables and notes in a PDF have to become rows and columns. With a spreadsheet, the supplier’s column names have to be understood.
  • Match values to your classes and fields. “Housing material” in one file and “Material (body)” in another may both belong to the same field in your Product class, or to a key in your classification store.
  • Convert units. The Data Importer’s Quantity Value operators pair a value with a unit id. The documentation describes no operator that converts between units. If a data sheet lists the weight as 0.09 kg and your field expects grams, someone has to turn it into 90 first.
  • Map select options. A supplier writes “Stainless steel”; your select field expects the code stainless_steel. Conditional Conversion can map known strings, but every new wording from a new supplier needs a new rule.

With one supplier and stable files, this is a one-time mapping job. With many suppliers whose files change every season, it becomes the daily work of the catalogue team.

Where Pimcore workflows help

Pimcore’s Workflow Management configures workflows on assets, documents and data objects. Places describe a status, transitions move an element between them, and guards decide whether a transition is allowed. A transition can require a note from the user, and a place can carry permissions, so a finished product cannot be changed by mistake. Workflows give imported products a review path inside Pimcore. They do not read supplier files or decide how a value maps.

Pimcore’s own AI and supplier files

Pimcore’s AI-driven platform page describes Pimcore Copilot, integration with many AI and ML models, zero-shot classification, computer vision and natural language processing. Here is what the documentation shows for onboarding work.

Pimcore Copilot. The Copilot documentation calls it the central starting point for running actions in Pimcore Studio, with a job engine for long-running jobs. Its included actions cover text generation, a variant generator, asset relation assignment, webhooks, filters and workflow transitions. The Hugging Face action steps add translation, text classification, zero-shot text and image classification, image to text and image upscaling. These help after products are in Pimcore. The documentation lists no action for importing supplier files or reading values from PDF data sheets.

Pimcore Agent. The Pimcore Agent Bundle, listed under Pimcore Labs in the docs, adds AI chat agents to Pimcore Studio “with human-in-the-loop write control”: every write goes through a propose, review and approve pipeline. Its MCP integration gives agents tools for data objects, assets, documents and tags. The installation page requires Pimcore 2026.x. The documentation mentions “data-import patterns” as an example skill, but describes no supplier file intake or PDF reading.

In short: Pimcore’s AI works on data that is already in Pimcore. Getting supplier files into the right shape is still a separate step.

Onboarding extensions in the Pimcore Store

Two listings in the Pimcore Store focus on supplier data onboarding.

Onedot Product Data Onboarding

The Onedot listing describes two jobs. The first is classification: Onedot AI classifies product data to standards such as ETIM and ECLASS, maps classes, features and lists of values, and your team switches to validating the AI’s work. The second is onboarding: supplier catalogues are sourced through Onedot and imported into Pimcore using your own data model. The listing names CSV, Excel, XML, JSON and text as catalogue formats, and lists normalisation of numbers, units and lists of values, variant grouping, golden records across sources, and asset onboarding for images, PDFs and videos. It requires a commercial Pimcore licence (Professional, Enterprise or PaaS) and Pimcore 2024.1 or later.

openProd.io AI Data Onboarding

openProd is listed in the Pimcore Store as openProd.io AI Data Onboarding. It is described in the next section.

How openProd works with Pimcore

openProd is an onboarding layer in front of Pimcore. Your team uploads supplier PDFs (product names and specification tables) and spreadsheets (price lists, update files), or pastes text from an email. openProd reads them and matches each value to an attribute of your own catalogue, not a generic template.

Your Pimcore model comes first. A connector job reads your Pimcore classes and fields and prepares matching families, attributes, units and dictionaries in openProd. When a mapping needs a business decision, the job ends with a question instead of creating something on its own.

Values land on your units and codes. A weight read as 0.09 kg is saved as 90 when the attribute expects grams. “Stainless steel” becomes stainless_steel, the code your dictionary uses. openProd suggests the product family and shows how many attributes match; nothing is set until you choose.

People decide what is uncertain. Clear values go straight through. Conversions and values without an attribute wait in Data review, where you Decide with the source one click away: Show in source opens the file at the place the value came from. You compare the value as read with the value after conversion, then save or skip it, and you can step back one decision. Values your model has no attribute for are listed separately, and deciding on them is optional.

Nothing is written by surprise. The preview separates new products from updates and shows each changed value, now and after import. Nothing is written until someone clicks Save to catalog. Before the AI analysis starts, openProd shows a credit estimate, and the AI runs only after you accept it.

Sync to Pimcore is checked. You pick products by family, status or search and sync them. Each product shows not synchronized, needs attention or synchronized, and it counts as synchronized only after the result is verified in Pimcore. Products that need attention say what went wrong and can be retried. Viewing the connector and running a sync are separate permissions. How openProd connects depends on your Pimcore version and installed components, and is agreed during setup.

openProd does not replace the Data Importer. Suppliers who send clean, stable files can stay on it. openProd covers the files that need reading and decisions first. LemonMind, the team behind openProd, is a Pimcore Platinum Partner. More on the Pimcore integration page and the AI import page.

Run a supplier onboarding trial with three real files

Whichever route you test, use the same three files so the results compare.

  1. Pick the files. One supplier PDF data sheet, one spreadsheet with the supplier’s own column names, and one product email or text block. Choose files your team found slow last quarter, not the clean ones.
  2. Write down the target. For each file, note the Pimcore class, the fields or classification store keys, the units and the select options the values should land on.
  3. Run the route. With the Data Importer, note what had to be prepared by hand before the file could be mapped. With an extension or onboarding layer, bring the files to the demo and run them live.
  4. Check the units and options. Did 0.09 kg arrive as 90 g where your field expects grams? Did “Stainless steel” land on your option code?
  5. Trace three values. Pick three values at random and find where each came from in the source file.
  6. Look at what is left over. What happened to values your model has no field for?
  7. Check what is written. What exactly reaches Pimcore, when, and who approved it? Can you see the change and undo it?

The tutorials show this flow in openProd, and in a demo one of your own supplier files goes through reading, mapping and review.

Which route fits your setup

  • Suppliers send structured files that rarely change: the Pimcore Data Importer, with a mapping per supplier and a schedule.
  • You need ETIM or ECLASS classification inside Pimcore: look at Onedot Product Data Onboarding in the Pimcore Store.
  • Suppliers send their own PDFs and spreadsheets, and you want every value checked against your Pimcore model before it is saved: an onboarding layer such as openProd in front of Pimcore.
  • Products are already in Pimcore and need copy, translation or classification: Pimcore Copilot actions, or an enrichment extension.
  • You want AI agents to work on Pimcore data under approval: the Pimcore Agent Bundle and its MCP integration.

For a wider view of onboarding tools beyond Pimcore, see our comparison of supplier product data onboarding tools.

Sources and updates

Checked on 13 October 2026. If something has changed, tell us and we will update this page.

FAQ

Questions and answers.

Can the Pimcore Data Importer import supplier PDFs?

The Pimcore Data Importer documentation lists six data sources (asset, upload, HTTP, SFTP, push endpoint and SQL) and five formats: CSV, JSON, XML, XLSX and SQL query results. PDF is not among them. To use a supplier PDF data sheet with the Pimcore Data Importer, someone first has to turn it into a structured file, by hand or with an onboarding tool.

Does the Pimcore Data Importer convert units?

The Pimcore Data Importer's transformation pipeline has Quantity Value operators that pair a value with a unit id. The documentation does not describe an operator that converts between units, such as kilograms to grams. If a supplier sends weights in kilograms and your Pimcore field expects grams, the conversion has to be settled before the value reaches the Data Importer.

Can Pimcore Copilot onboard supplier files?

Pimcore Copilot's documented actions cover text generation, translation, text and image classification, image to text, variant generation, asset relations and workflow transitions. The Pimcore Copilot action list does not include an action for importing supplier files or reading values from PDF data sheets. Check the Copilot documentation for your Pimcore version, because the action list can change between versions.

Which Pimcore Store extensions help with supplier data onboarding?

Two Pimcore Store listings focus on supplier data onboarding. Onedot Product Data Onboarding classifies products to ETIM and ECLASS and imports supplier catalogues into Pimcore using your data model. openProd.io AI Data Onboarding reads supplier files and maps every value to your own Pimcore classes and fields. Check each Pimcore Store listing for edition and version requirements.

When is the Pimcore Data Importer enough for supplier data?

The Pimcore Data Importer is enough when a supplier sends structured files (CSV, XLSX, XML or JSON) with stable columns, units and option values that already match your classes. You set up the mapping once, check it in the live preview and run it on a schedule. It needs help when files arrive as PDFs, change layout or use other units.

How do I test supplier data onboarding for Pimcore?

Take three real files from your suppliers: a PDF data sheet, a spreadsheet with the supplier's own column names and a product email. Run them through each Pimcore onboarding route you consider. Check whether values land on your Pimcore classes, units and option codes, whether you can trace each value to its source, and what is written before you approve it.

More from the Academy.

See it working on your own files.

Book a short demo and we will run a real supplier file through openProd with you.

Book a demo All articles