Microsoft Launches SharePoint Syntex To Automate Content Categorization And Build A Foundation For Knowledge Curation
Today, Microsoft announced the general availability of Microsoft SharePoint Syntex as of October 1, 2020. This is the first packaged product to come out of the code-name Project Cortex initiative first announced in November 2019. Project Cortex reflects Microsoft’s ongoing investment in intelligent content services and graph APIs to proactively explore and categorize digital assets from Microsoft 365 and other connected sources.
SharePoint Syntex will be available to M365 customers with E3 or E5 licenses for a small per-user uplift. As of this writing, we anticipate it to be around a $5 per-user per-month list price, but this may be subject to change. SharePoint Syntex delivers some of the foundational artificial intelligence and machine-learning (ML) services that will help information managers understand, process, and tag content automatically. The second phase of the Project Cortex launch — tools for knowledge curation and management — is expected in later 2020.
Why Is This Important?
Too many organizations have ignored the importance of a solid information architecture and metadata strategy — whether they are using SharePoint or not. The enhancements delivered in SharePoint Syntex could help get these strategies back on track. Organizing and tagging documents at large scale is a daunting task that currently requires a great deal of human labor — but is important work to form a strong foundation on which to build an enhanced set of knowledge discovery, delivery, and curation capabilities in the near to mid-future. Sharepoint Syntex is positioned to automate some of this intensive labor and drive key outcomes.
What Is It, And What Does It Do?
Microsoft SharePoint Syntex will deliver new ways of managing large volumes of documents via a new “content center,” which brings various intelligent content services — AI, ML, optical character recognition, enhanced taxonomy services, etc. — to document libraries. Microsoft is taking some of the most relevant Azure cognitive services and infusing them into M365 via this SharePoint Syntex add-on product. New model-building features will allow subject matter experts and content stewards to define and refine how the intelligent services analyze, tag, and extract data from documents.
Highlights of SharePoint Syntex available in October include:
- Image and forms processing. Images can be automatically tagged by leveraging what Microsoft calls a “new visual dictionary” to apply metadata descriptors when common objects are recognized in an image (including JPGs, PNGs, PDFs, and so on). Another service allows nontechnical users to build an AI model to automatically extract values from semistructured documents, such as dates, names, or addresses, from repeatable document types such as receipts or invoices. Microsoft claims that these form processing models can be trained with a small set of sample documents — perhaps fewer than 10 — if the right mix of positives and negatives are included.
- Document understanding. Longer text-heavy documents may have broad or long-term business value and benefit from consistent metadata tagging for better search and discovery. SharePoint Syntex can automate metadata tagging of content-rich documents. Microsoft has built this capability using the Language Understanding Intelligent Services for Documents (LUIS-D) model from its Azure Cognitive Services. These models, also built in the new content center, are trainable by subject matter experts and can be applied to multiple libraries. Formats include Office documents, text formats, PDFs, emails, etc.
- Automated compliance labels. This automatically extracted metadata not only aids in better search and retrieval, but it can be used to initiate a workflow process, apply a retention policy via Microsoft’s newish retention label feature, or leverage sensitivity labels to control access and distribution of the document.
What Can Organizations Do With Microsoft SharePoint Syntex?
Microsoft customers can work toward automating the organizing and tagging of documents (at scale) by:
- Experimenting with Syntex with a subset of your user licenses. This is not a feature to be flipped on and just work. It will require investment in time and internal expertise. Organizations wanting to pilot SharePoint Syntex can start with a small set of add-on licenses to get things started. Pick a set of documents or use cases that are causing productivity bottlenecks, are part of integral processes that can be driven by metadata, or that can enhance adoption of related retention or data protection policies with more consistent tagging. The initial release of SharePoint Syntex will support English, with other languages to come in the future.
- Working with a specially trained partner to hit the ground running. To Microsoft’s credit, it is not positioning these Cortex-inspired products as technology magic bullets. To make SharePoint Syntex (and subsequent product releases) really work, it will require human expertise, knowledge of business processes and information architecture, and skills to get projects up and running. Microsoft has launched a partner program specifically to train and enable select system integrators and independent software vendors, which can then support end-user customers.
- Gathering your information and knowledge management gurus into a dream team. Bring your experts to the (virtual) table and understand where to invest next. In conversations with large enterprises over the last 18–24 months, it is clear to me that there is a renewed interest in managing digital knowledge assets better and smarter. The need to support virtual and remote workers has upped the stakes on a solid strategy for information management. Companies that will survive — even thrive — in the tumult that is 2020 understand the value of knowledge to serve customers as well as employees.
Want to have a conversation about Microsoft SharePoint, M365, or the broader content management landscape? Contact us to set up an inquiry.
(J. P. Gownder, vice president, principal analyst, contributed to this blog.)