Solutions
Synopps software: one engine, three modules for master data tasks
Academy
Master data training built on real project practice
The requester describes the item in a chat, in their own words. Synopps runs the checks along the way, asks only for what is missing and hands a finished record to the master data specialist, who confirms it.
Synopps New Record is a tool for creating new items in the material master through a chat. It sits where the record is born: the requester describes the item in their own words, and the tool checks the description, searches for duplicates across the whole material master, proposes a class and template, extracts attributes from the text and confirms from web sources that the item exists. It suits companies whose material master is already large but who have no dedicated master data team, and it works as an add-on where an MDM is already in place.
The requester does not need to know the classification, remember the naming convention or keep in mind what is already in the material master.
The class and template are matched to the meaning of the description. The chat shows five close options with a similarity score, and a person picks one.
Dimensions, materials, ratings and the part number are extracted from what the requester wrote. The short description is built with the single naming convention of your material master.
Instead of ten open browser tabs, the tool goes through the search results and brings back the sources, showing what each one confirmed.
The checks start as soon as the description is sent. Class, attributes and web search run when a person asks, and no value is filled in by guesswork. The tool prepares, the master data specialist confirms.
Each of these situations starts with one item that nobody checked when it was created.
A one-off cleansing of the material master does not solve these situations: it fixes what has already piled up, and it does not change how the next item gets created.
The check comes after the request. Duplicates are not always found.
Duplicates are visible before the item is created. The expert's decision stays with the record.
The transfer to your ERP (SAP or another system) is set up during the project for your landscape.
The requester does not face a form with dozens of fields. They write what they need, and the chat leads them through the same seven stages a master data expert would follow.
The requester describes the item the way they would to a colleague. No codes, no class tree, no naming rules to learn.
What is already in the description is filled in. The chat comes back only for the fields the material master still needs.
Class, names, duplicates and sources arrive as cards with buttons. Nothing is saved until a person chooses.
As soon as the description is sent, it is checked for double spaces, forbidden and non-printing characters, mixed character sets, unbalanced brackets and descriptions that are too generic. What can be fixed is fixed, and the chat says what changed.
The item is checked against the whole material master. Similar items appear in the chat with a score and the method that found them. If there are any, the conversation does not move on until a person answers: duplicate, not a duplicate, or needs review.
Five classes with a similarity score, matched to the meaning of the description. The requester picks one with a click, and the attribute template comes with it.
Values are read from the description into the template. The short and long names are built with your naming convention, and the short name is held to the length limit of your ERP.
The manufacturer and part number pair is checked against web search results. Each source shows what it confirmed. The step runs only when someone asks for it.
Project fields beyond the template, such as unit of measure or MRP type. What to ask is configured for you. Mandatory fields hold the conversation until they are filled in.
One summary of the item and every check. The requester submits it for approval, and the master data specialist decides whether it reaches the material master.
The stage bar stays above the conversation, so it is always clear what is done and what is left. An item cannot be submitted until every stage is closed.
The tool proposes, the master data specialist accepts. An external AI agent connected to the tool cannot overturn a person's decision.
A duplicate shows the matching text and the method that found it. A class shows five candidates with similarity scores, not one verdict.
Runs, pairs, decisions and exports of past versions stay as they are. A new run does not overwrite an accepted result.
A value that is not in the data is never filled in. An empty field stays empty, and a failed check is called failed.
The tool has one job: the entry point to the material master. Whether you have an MDM only changes where the finished item goes.
The tool controls the entry point on its own. Naming rules, classification and attribute templates are set up for your material master. You do not need to create a separate master data team. The finished item goes to your ERP.
Storage, distribution and the master data architecture stay in the MDM. The tool takes the narrow job at the entry point: check and build the item before it appears in the material master. If you still need to implement an MDM, there is a separate page on that.
This is a case where a light layer at the entry point solves a problem that usually calls for a full master data management programme.
Exchange runs through an API, so the ERP can be any: SAP or another system. The tool is also open to agents through its own MCP server.
The finished item goes to SAP or another ERP. Exchange runs through an API and is set up for your landscape during the project.
The tool has its own MCP server. An external agent connects to it and runs cleansing and duplicate checks across the whole material master. If you only need the engine inside your own system, that is Synopps Core.
Connect a cloud model of your choice or a local model inside your perimeter.
A connected agent prepares and proposes. It cannot overturn the master data specialist's decision, and its actions stay in the item history.
A short, honest list. It saves time in the very first conversation.
We start with a part of the material master where the result is visible, and with one category of requests.
We look at your material master, the request categories and how items are created today.
Which part of the material master, which categories, what counts as a result and where the finished item goes.
Naming rules, classification and templates are set up for your material master, requesters create items in the chat, and the expert confirms.
A form expects the requester to know the classification and the naming rules in advance. In the chat they describe the item the way they would to a colleague, and the tool asks only for what the material master still needs. The structure behind it stays strict: seven stages, the same checks and the same approval.
No. The tool is built for companies whose material master is already large but who have no separate process or team around it. Items are created by the same people who create them now. A master data specialist confirms them, and as rules build up, that work comes down to confirmation.
Your choice: in the browser on our side, or installed in your environment. In the second case the master data never leaves your network.
We do, together with your experts. Classification, attribute templates and naming conventions are built for your material master at the start of the pilot.
The material master is loaded into the tool and serves as the basis for the duplicate check. Cleansing what has already piled up is a separate job, and we do that too.
Yes. Exchange runs through an API, so the ERP can be any: SAP or another system. You can also connect an external AI agent to the tool through the MCP server.
No. Without an MDM, the tool controls the entry point to the material master on its own, and the finished item goes to your ERP. With an MDM, the tool works as an add-on: storage, distribution and the master data architecture stay in the MDM, and the tool takes the job at the entry point.
The new item is checked against the material master as soon as the description is sent. Similar items are shown in the chat with a score and the method and text fragment that produced the match. A person makes the decision, and it is saved together with its basis.
Cleansing fixes what has already piled up and does not change how the next item gets created. The tool controls the entry point itself, so a year later the material master has not slipped back to where it was.
Tell us about your material master and how items are created today. Then we agree the scope of a pilot.