Master data management, automated: article master data that stays current, complete, and duplicate-free on its own.
Every system, every channel, one reliable record per article: Syntact automates the ongoing maintenance — reads every update, maps it to your article model, checks for duplicates against your existing data, and puts everything up for approval. Your team approves — instead of typing.
Reality changed a while ago. Your article master still doesn't know.
Not because your data team is slow — but because between the sources and your article master sits a media break that people have been bridging by hand. Every update, from every source, next to the day job.
The ERP says A, the shop says B, the Excel file in sales says C. Nobody knows which record is right — so people call around, look things up, and when in doubt create the article again. There is no single source of truth. Just three versions of one.
Every new channel exposes the gaps: the shop needs descriptions, the marketplace demands mandatory attributes, the catalog needs clean units — and tomorrow the digital product passport will ask for even more. The master keeps up with none of it.
Created once via the EAN, once via the supplier article number, once by hand: without duplicate checking there is no golden record — and every report, every order, and every price depends on which of the three records someone happens to hit.
Every update — price list, catalog, datasheet — is read, mapped, and entered by hand. Article by article, next to the day job. The backlog grows faster than the team can type.
A wrong master data record never costs just one typo.
Manual master data maintenance has two built-in costs: the backlog that never gets caught up — and the silent errors it regularly writes into your article master. Both grow with every update.
A shifted decimal, a mixed-up unit — or the update simply sits in the backlog. Neither gets noticed. The article carries the wrong state into every system that reads the master.
A wrong record doesn't trigger an error message. It just keeps running: wrong orders, returns, negative margin — on every order, in every channel. Master data errors are the most expensive kind, because nobody reports them.
Weeks pass between 'update received' and 'update entered.' During that time, every system works on data that stopped being true a while ago — purchasing, sales, the shop, every report.
The same article created three times, attributes sometimes maintained, sometimes empty, units depending on who typed them: duplicates and gaps make every report untrustworthy — and the dataset gets worse with every unchecked update.
What does master data maintenance cost you — the cost nobody questions?
A conservative calculation: pure maintenance time only. The silent follow-up costs — wrong orders, returns, reports on stale data — and the cleanup of duplicates aren't even counted yet.
Approve updates instead of entering them.
No supplier portal, no mandatory format, no rebuilding your sources: updates keep arriving the way they arrive today. Only the data entry is taken over by AI — embedded in your existing system landscape.
An Excel price list, a PDF catalog, a datasheet, an ERP export, or the attachment in a supplier email: Syntact monitors the intake continuously — nobody converts anything, nobody builds mapping files. The 40-line update just like the 40,000-article catalog.
The AI identifies articles, attributes, units, and variants — and maps everything to your article model. Every record is matched against your existing data via duplicate checking: existing articles are updated instead of created again, and missing attributes are enriched with their source.
Your team sees the finished update as a proposal: changes, new articles, flagged exceptions — next to the original source. Approve, and the golden record in your ERP and PIM is current again. Every system reads the same truth — weeks of typing become minutes of control.
Approval stays with your team: no update lands in the master without the okay of your data owners. The AI proposes — people decide.
MDM systems give master data a home. Syntact fills it — and keeps it current.
The home doesn't maintain itself: every update from every source has to be read, mapped, deduplicated, enriched, and approved. That is exactly the layer Syntact automates — the approved golden record in your ERP/PIM is the goal, everything before that is an intermediate step.
Excel in any column logic, PDF catalogs, datasheets, Datanorm, BMEcat, attachments straight from email: Syntact reads whatever arrives. EDI stays where it pays off — for the 200 sources without EDI, Syntact takes over.
Everything is mapped to your product groups, attributes, and units — not to a standard schema that someone has to bend back into shape afterwards. Your data logic stays the reference, not ours.
Every article is checked against your existing data: EAN, supplier article number, description. Existing articles are updated instead of duplicated — three records become one golden record, and the master gets cleaner with every update.
Syntact fills missing attributes from catalogs and other supplier documents — and shows where every value came from. No silent guessing: every enrichment stays traceable.
The AI proposes — people approve. An unclear unit, a suspicious price jump, a possible duplicate: flagged and put in front of your team for a decision, with the original source right next to it.
Your ERP/PIM stays in the lead — Syntact is the maintenance automation behind it, not another data silo. If you run Pimcore or Akeneo, this makes them more valuable, not redundant.
Your ERP/PIM stays. Syntact keeps it current.
Updates come from the inbox and your source systems, verified records land in your ERP and PIM — via API or standard interface, with your article master as the reference. No system migration, no additional data silo.
We build missing connectors for free — contractually guaranteed. Compatible with anything that has an API or an import format for article data.
Master Data Pilot: your 3 most labor-intensive data sources — automated into your article master in 14 days.
A pilot on your real data intake, with your article model and your team — on terms where doing nothing is the most expensive option.
If your product data pipeline isn't running in production after 90 days, we keep working for free until it is.
You pay no platform license as long as the pilot isn't in production. The risk sits with us.
Catalog intake, article model mapping, and your approval workflow — set up for you, not sold to you.
90 days of direct access to the person building your use case — no ticket system, no call center.
If the pilot doesn't go to production, you pay only half the implementation.
Missing an interface to your system? We build it for free — contractually guaranteed.
The mechanism is already running — at real customers, with real data.
This exact use case runs in production at McSHARK: product, pricing, and logistics data from the most diverse sources, harmonized in one semantic data layer — with ROI from the very first use case. Syntact is the AI foundation platform for enterprises: one foundation for every use case, operated in the EU, all the way to on-premises.
- Supplier sources, catalogs, and the article master in one semantic data layer
- Permissions & audit log at every level
- After master data: order intake, invoice checking, knowledge assistant — same platform
"With Syntact, we successfully unified all our product, pricing, and logistics data and achieved ROI from the very first use case."
First, see how it works live. Then: your demo, on your data.
Two steps, zero risk: a 30-minute call with a live demo first — then, if it fits, a custom demo we prepare for free on your data.
- 1Book today: 30 minutesLive demo on a real example: A supplier update becomes a verified golden record in your article master. We capture your requirements — and tell you honestly whether it's a fit.
- 2Then: your demo. Your data.We prepare a custom demo with your documents, free of charge — ready in one week, presented in 60 minutes.
- Founder & CEO of CodeFlügel — software & AI from Graz since 2011
- 15 years of AI, app and data projects for industry and trade
- Projects for companies like Siemens, Henkel and voestalpine
In the call you talk directly to an expert — no sales call center.
- What does automated product data management mean?
- Every inbound update — no matter the source or format — is read and structured by AI: articles, attributes, units, variants. The data is mapped to your article model, matched against your article master via duplicate checking, and written to your ERP and PIM for approval. Your team reviews and approves — one golden record per article, continuously current, without typing.
- We're really just looking for a way to import price lists. Is this too big for us?
- No — price lists are the sharpest entry case, and there's a dedicated page for exactly that: 'Price list automation' at syntact.io/price-list-automation. This page is for the team that owns the article master as a whole: quality, completeness, duplicates, and ongoing upkeep across every inbound source. Many start there and expand — the mechanism is the same.
- Which sources can Syntact pull master data from?
- Practically everything that arrives today: Excel and PDF files, supplier catalogs, datasheets, ERP exports, email attachments — plus structured formats like Datanorm, BMEcat, or EDI messages, including the cases where suppliers interpret the standard creatively. The biggest lever sits with the sources that have no import path at all today: every update is read, mapped, and put up for approval — without a single supplier having to change their format.
- We have interfaces and EDI between our systems. Why Syntact then?
- Keep both — interfaces transport data, but they don't check quality. Duplicates, gaps, and conflicting states simply travel through every interface unchanged. Syntact sits in front: it reads inbound updates, checks every record against your existing data, fills missing attributes with their source, and puts everything up for approval. Only then does one truth propagate across your systems — instead of three versions.
- Our article model is special — variants, tiered prices, custom attributes. Does that fit?
- That is exactly what the mapping is for. During the pilot we define your target model together: product groups, attribute sets, unit logic, variant structure. Syntact maps every inbound update to that model — not to a generic schema. Anything that can't be matched unambiguously is put in front of your team for a decision instead of guessed.
- What is a golden record — and how do we get there?
- A golden record is the one reliable dataset per article: current cost price and RRP, verified attributes, one unit logic, no duplicates. Syntact doesn't build it in a big-bang project, but continuously: every update is matched against your existing data via duplicate checking, conflicts are put in front of your team for a decision, and every approval makes the master cleaner.
- What happens with duplicates and articles that already exist?
- Every article is checked against your existing data — via EAN, supplier article number, and description matching. Existing articles are updated, for example with new prices, new ones are created, and uncertain cases land flagged in the review queue. The master grows in a controlled way instead of wild — duplicate checking included, on every single update.
- Honestly, our data is a mess. Do we need to clean up first?
- No — the cleanup is part of the mechanism. Matching against the article master surfaces duplicates and gaps, enrichment fills missing attributes with their source, and every approval improves the dataset. Whoever waits for clean data before automating waits forever. It works exactly the other way around.
- How is this different from a PIM project?
- A PIM project gives your product data management a better home — but it doesn't fill itself and doesn't keep itself current. Syntact doesn't replace a PIM and doesn't introduce one: your ERP or PIM stays in the lead, and Syntact automates the ongoing supply — from the inbound update to the approved record. If you already run Akeneo or Pimcore, this makes them more valuable, not redundant.
- We're currently evaluating PIM/MDM systems (e.g. Pimcore, Akeneo). When does Syntact make sense?
- Before, during, or after the rollout. The PIM/MDM is the home of your master data — Syntact is the steward that fills it and keeps it current: reading updates, mapping, checking duplicates, enriching, putting everything up for approval. Combining both saves you the data maintenance team, not the system. And if you don't have a PIM yet, you start with Syntact directly on the ERP — the home can be added later.
- Day-to-day business is running at full capacity — we have no time for a project.
- The pilot is built to run alongside your day-to-day business: our solutions architect sets up the pipeline, your team provides one typical data source, a master data extract, and feedback in short sessions. And the best moment is right now — the update sitting in your inbox today will otherwise be typed in by hand again.
- Who do I need internally — and what do I tell management and IT?
- For management: one calculation — maintenance hours per year times hourly rate, plus the silent follow-up costs of duplicates and outdated records, against a license from €800 per month that only starts once you're in production. For IT: little effort — a data intake (e.g. a mailbox), an interface or import format for your ERP/PIM, and a master data extract. Our solutions architect does the setup. And for both, there's the demo on your real data.
- What about data protection and our supplier data?
- Syntact runs GDPR-compliant in the EU — in your own cloud or on-premises if you prefer. Article, condition, and supplier data stay in your systems, and access follows your permission model. We provide the documentation your data protection officer and purchasing department need.
- What does automated product data management cost?
- The platform starts at €800 per month — and the license only starts once you're in production. Setup, data intake, and the ERP/PIM connection are included in the pilot package, and implementation is covered by the 50/50 guarantee. A single update round with a few thousand positions costs more to enter manually than the platform costs in a whole year.
One inbound channel hurts the most? Start exactly there.
Price lists and supplier catalogs are the most labor-intensive sources feeding your article master — each has its own page. And the same mechanism reads your order intake, too.
The next update is definitely coming. The question is who enters it.
Minutes of review instead of weeks of typing — and an article master every system and every channel can trust.