Improve your Documents and Quality Management Systems with Microsoft Syntex

Whilst organisations have been basing their document and/or quality management systems (DMSs and QMSs) in SharePoint for years now, there has quite often been an administrative overhead in maintaining them. This is particularly relevant when users are required to manually tag files with custom metadata values.

Though SharePoint enables you to heavily customise and personalise metadata, it’s still proved tricky to automate much of this through default column values, calculations and required fields. There is still too much reliance on users to manually tag, as well as susceptibility to human error. Of course, there is unlikely to be a point where no user involvement is required (at least anytime soon!), but at least in the SharePoint space, Microsoft Syntex is etching ever closer.

What is Microsoft Syntex?

Microsoft Syntex uses AI to manage and organise content, optimise search and compliance, to automate and improve your most critical business processes.

Microsoft Syntex enhances SharePoint’s capabilities by automating the process of classifying and extracting key information from documents. This is done by configuring Microsoft Syntex no code AI models for processing documents. These models can then be trained to understand key information, extract metadata, apply retention/sensitivity labels and secure the content, so no manual data entries or data tagging is required.

Microsoft Syntex is now much more than document processing and has very exciting roadmap for new functionality. To find out more about these and Syntex’s full capabilities, please see the following article: Microsoft Syntex launched at Ignite.

How can Microsoft Syntex improve our Document/Quality Management Systems?

In the context of document and quality management systems, Microsoft Syntex’s extraction tooling is a real game changer. Imagine being able to automatically extract the key information from a document and tag the document with it as metadata columns, without even the click of a button. Sounds too good to be true right? With the right Microsoft Syntex model applied to your document library though, this is entirely possible. Welcome to what Microsoft call ‘entity extractors’.

Entity extractors are used as part of a Microsoft Syntex model to basically do what you would expect, extract entities from your documents. Whether you’re dealing with client names, contract dates, profit totals, addresses or phone numbers, Microsoft Syntex can automatically extract these from you documents and set them as metadata values. This automatic extraction of metadata is not only automating a typically inconvenient and timely process but it is also improving the quality and findability of your information.

Once extracted and populated in your metadata columns, your documents are now filterable, sortable and key information is now far more visible at significantly reduced administrative overhead.

Microsoft Syntex Example

The example below shows a simple quality management system used by a company to manage its policies, including metadata fields such as ‘Owner’, ‘Department’ and review dates.

Without a Microsoft Syntex model, much of the required metadata information needs to be manually populated by the user. Certain defaults such as Content Type and Retention Label could have been applied by default though, this is not always appropriate of course.

In contrast, below is the same library with a simple Microsoft Syntex entity extraction model applied to it. There has been no manual tagging completed to collect this metadata, this is all through Microsoft Syntex:

So, what is actually happening here? Firstly, let’s take a closer look at one of the documents in question.

In this ‘Annual Leave Policy’ example we have highlighted the values that we require for the library columns, despite some actually being part of the paragraph text (department and review dates). By using Microsoft Syntex’s capabilities, the applied model is trained to extract these values automatically from the document and populate the library columns.

However, the model is also recognising this as a ‘QMS Policy Document’ content type, and so appropriately applies that information to the document also. This enables you to apply multiple trained models to the same library to identify and tag different types of documents.

How to create and deploy a simple Syntex model

In contrast to other Microsoft AI or Machine Learning offerings, Microsoft Syntex is a far more friendly and UI-configurable service. Let’s go through how you might go about incorporating Microsoft Syntex with your SharePoint quality/document management system using the example above.

CREATING A MODEL

Step 1: Head to your Content Centre SharePoint site, navigate to the Models tab, and select ‘Create a model’.

Step 2: Select the ‘Teaching method’ model (also known as unstructured). This offers us the best options for the QMS style documents which are relatively unstructured.

Step 3: The next page in the dialog box will provide you with some insight on what the model can do and where you might use it. For now, click ‘Next’ to proceed.

Step 4: Enter a name for your model and select whether you would like to use an existing content type (which you may already be using in your QMS/DMS) or to create a new one. In this example, we’re going to create a new content type.

TRAINING YOUR SYNTEX MODEL

Step 5: Select ‘Add files’ to upload your training files and then add them to your ‘Example files’.

Step 6: …

Download the FREE guide below for the next steps. This guide has been carefully prepared by our Microsoft 365 consultant with detailed steps and supporting images.

Other options and features to consider

Though powerful, the Microsoft Syntex model described here is reasonably simple. There are numerous other features the service can offer that you might use for a quality/document management system, such as:

• Document Summarisation (Coming Soon)
• Document translation (Coming Soon)
• Document annotation
• Content assembly

Content assembly is a very exciting option you might want to explore, especially if your quality/document management system is already using templates. Content assembly will utilise Modern Templates to assist in creating documents using a form. This will not only save you time creating new files, but will also improve the consistency of your information.

Although only briefly touched upon, Microsoft Syntex’s ability to apply retention labels to content dynamically through its modelling is also something we are really excited about. Yes, Microsoft Purview has features such as Trainable Classifiers which you can use to adaptively apply to content, but the Microsoft Syntex usability and UI take this too a whole new level.

PS: Do you want us to do all this for you or do you need help deploying Syntex, give us a call.