Why we built openProd

Introduction

Across more than 70 PIM and e-commerce implementations we kept seeing the same problem. It delayed project schedules, stretched budgets and frustrated every team: getting supplier data into the PIM. Not the PIM itself. Not syndication to channels. Not the DAM. Data onboarding. That is why we built openProd, an AI-native platform for product data onboarding, designed to remove this bottleneck. openProd is a separate product with its own site, openprod.io.

The cost nobody puts on the slide

Few PIM presentations show how much of a project goes into onboarding supplier data: extracting data from supplier catalogues, mapping fields, normalising values, checking quality and fixing errors. In PIM projects the most time goes not into the system, but into this work.

We know this from experience

We have done it many times, for clients in manufacturing, retail and distribution.

A typical supplier sends data as a PDF catalogue, an Excel file with dozens of tabs or, if you are lucky, a CSV that is almost usable. The data team spends a long time on every supplier, cleaning, mapping and loading. Multiply that by the number of suppliers and the PIM schedule grows quickly.

That is the hidden problem of PIM projects. The technology works. The data does not.

Why traditional approaches fail

Since 2010 we have tried everything:

Manual data entry teams. Expensive, slow and error-prone. It does not scale.

ETL scripts and custom importers. Great for structured data. Useless when a supplier sends a long PDF catalogue with nested tables.

Outsourced data services. Black-box quality, long turnaround and no knowledge kept inside the company.

"AI" from PIM vendors. Mostly cosmetic: generating descriptions or translating text. It does not touch the hard problem, which is extracting structured data from unstructured files.

None of these solve the core issue: turning chaotic, inconsistent supplier files into clean, validated product data at scale.

What openProd does

openProd is built for one job: moving product data from supplier files into your PIM, quickly and accurately.

1. Upload a supplier file. PDF catalogues, Excel sheets, CSV, JSON and XML files, and images. No preprocessing and no templates.
2. AI extracts and maps the data. The extraction engine identifies product attributes, maps them to your PIM schema, normalises values (units, currencies, naming conventions) and generates missing descriptions.
3. A confidence score on every field. Each extracted value gets a confidence score. You approve high-confidence data in bulk, and low-confidence data goes to manual review. Your time goes where it matters.
4. A cost estimate before you run it. Before any extraction starts, you see the estimated token usage and cost. No surprises and no runaway API bills.
5. Re-Prompt Engine. Found the same error in many products? Describe the fix in plain language, for example "convert all weights from kg to grams", and the AI applies it to every affected product.
6. Push to the PIM. Direct integration with Pimcore, Akeneo, Ergonode or any other system through a REST API.

MCP server: why it matters for the future

openProd includes an MCP server (Model Context Protocol): a standard interface that lets AI agents work with your product catalogue directly.

Why does it matter? Because enterprise software is becoming agentic. AI agents will need to read, query and update product data inside automated workflows. MCP is an emerging open standard for exactly that.

With the openProd MCP server, your product catalogue is ready for that future today.

Read more on the openProd blog.

PIM-agnostic by design

A key architecture decision: openProd does not replace your PIM. It feeds it.

At LemonMind we are a Pimcore Platinum Partner, we implement Akeneo and we work with Ergonode. We have seen what happens when companies depend on one vendor's ecosystem for data onboarding: they lose flexibility, negotiating power and the option to switch.

openProd is PIM-agnostic. It works with whichever PIM you have now or will have later. The onboarding pipeline should not be tied to the catalogue system. That is a lesson from more than 70 implementations.

Who is openProd for?

• Product data managers drowning in supplier spreadsheets
• PIM managers who need a faster time to market for new products
• Heads of e-commerce fighting data quality problems in sales channels
• CTOs looking for an AI-native approach to what is, at heart, a data problem
• CFOs who want a defensible business case with measurable ROI

If your team spends more time cleaning data than managing products, openProd was built for you.

What next?

When we published this article, we were onboarding the first customers and developing the platform based on real use. If you want to see openProd working on your own supplier data, book a demo on openprod.io.

If you already work with LemonMind on Pimcore, Akeneo or Ergonode, ask us about connecting openProd to your current PIM workflow. The onboarding bottleneck does not have to be yours.