The blog post by Leon Armston reports on the significant updates to Microsoft Syntex, also referred as the Structured Document Processing model.
The most valuable upgrade allows models to be replicated across libraries and sites, eliminating the need to set up unique models for each library. This offers convenience in deploying structured document handling models to various libraries/sites.
The model leverages Microsoft Power Apps AI Builder document processing, which uses machine learning to pinpoint and extract key-value pairs and table data from documents. Both the AI Builder team and Syntex released this novel integration enhancing its capability and efficiency.
A wealth of improvements have been made to the model creation experience and related UI. The design offers a central platform for creating all Syntex models from a single menu. No longer does it rely on Power Apps AI Builder screens or waiting for environments to load.
The internal workflow service now handles all processes, eliminating the Power Automate dependency. While this change comes with minor limitations, it simplifies the model creation process significantly.
The author then guides readers through creating a Structured Document Processing model via the integrated UI, and how to publish this model on multiple libraries. The instructions are easy to follow, outlining each step from naming the model to its training and deployment.
Another update allows Structured Document Processing Models to be published in bulk using PnP PowerShell, offering numerous automation choices. However, it is currently not possible to create a template for your structured dpm or freeform dpm and move it to another tenant or Syntex content centre using PnP PowerShell cmdlets.
In sum, the updates for Structured Document Processing Models have significantly improved its usability, especially the single-pane model creation UI and the ability to deploy models on multiple libraries and sites. Click here for more.
Microsoft Syntex is a powerful tool that uses machine learning to assist businesses in managing their data more efficiently. It can automate complex processes, extract important information from documents, and make data-driven decisions easier. The new updates significantly enhance its functionality, making it a valuable asset for any business that deals with large amounts of structured or semi-structured data, such as forms and invoices.
Microsoft's AI-based model, Microsoft Syntex, is a powerful tool designed to extract valuable information from structured or semi-structured documents such as forms and invoices. What makes it unique is its ability to learn, enabling it to recognize patterns in your documents and make more accurate predictions over time.
This ingenious technology recently witnessed significant updates, enhancing its capabilities and efficiency even more. Now, you can effortlessly use the same model across various libraries and sites. Earlier, you had to set up each model individually on a single library with no option to deploy the same model on other sites.
Syntex's structured document processing now supports Microsoft Power Apps AI Builder for identifying and extracting key-value pairs and table data. Additionally, the integration between Syntex and AI Builder has seen incredible development, leading to a more advanced and powerful, native integration along with the introduction of a new model type called freeform.
Considerable improvements have been made to the creation experience of the Syntex models and the associated user interface. Creating a Structured Processing model is now as easy as selecting the layout method button from the new model creation UI.
Previously, a structured document processing model, once created and trained in AI Builder, initiated a flow in Power Automate to read and update documents. However, with the latest improvements, the dependency on Power Automate is eliminated. An internal workflow service now does the reading and updating of files without the need for Power Automate.
Creating a model with the layout method is a straightforward process. First, you need to specify the model's name and provide a description. Following this, you have to specify the fields that you wish the AI model to extract and identify their type. By uploading at least five sample documents for the model to study, you assist the model in learning. Different documents with varied layouts can be grouped into collections.
Sections of the document are mapped to the fields specified earlier for extraction. After clicking on a piece of text, you assign it to a specific field. When the model creation and modification process is completed successfully, a model summary page is displayed for review. If everything appears to be in order, you can send the model for training.
A review of the created model can be conducted, and information about the extracted data and its accuracy can be evaluated. Once you are satisfied with its performance, the model can be published and applied to a library.
The structured document processing models can be applied in bulk using PnP PowerShell. But remember, the PnP cmdlets work only with the Unstructured Document Processing models.
Furthermore, it must be clarified that creating a model with structured document processing model technology or freeform technology and moving it to another tenant is not possible. The reason behind this is that creating such models involve storing the model configuration in Dataverse tables, powered by AI Builder.
Overall, Microsoft has made some very exciting updates to its structured document processing model, enhancing its convenience and efficiency, especially the models' re-usability and the integrated Syntex model creation interface in a single pane.
Microsoft Syntex updates, Document Processing Model, Important Microsoft updates, Structured Document Processing, Syntex model updates, Microsoft Structured Document, Updates Microsoft Syntex, Structured Document Model, Microsoft Syntex important updates, Microsoft Document Processing.