Microsoft launched Project Cortex at Ignite 2019, emphasizing its utility in data discovery and knowledge management as part of the data services in Microsoft 365. Project Cortex’s AI algorithms are used to search through company data to identify the resources around key products, services, and people associated with those knowledge areas. That information is accessible through corporate search tools to create opportunities for cross-organization synergy and cooperation.
That approach has provided the incubation and proving ground for SharePoint Syntex, the first Microsoft 365 service built directly from Project Cortex. SharePoint Syntex lets customers train the service to read a document and extract information, then uses customer models to automatically process content, extract information and apply metadata.
According to Microsoft, SharePoint Syntext focuses on three distinct areas:
- Content Understanding: AI models are trained to classify and pull out information from documents. This information is then added to the documents’ metadata to facilitate better discovery in search queries.
- Content Processing: Using Power Automate, processes are built to automate the capturing, consumption, and categorization of content across the company.
- Content Compliance: Project Cortex can apply company data protection policies. This ensures business-critical data receives the proper level of protection throughout its handling by members of the organization.
Pay attention to the term machine teaching versus machine learning? Microsoft differentiates between these two methods for SharePoint Syntex:
Machine teaching accelerates the creation of AI models by acquiring knowledge from people rather than from large datasets alone. Any information processing skill, that an expert can teach a human, should be easily teachable to a machine.
SharePoint Syntex mainstreams machine teaching, enabling your experts to capture their knowledge about content in AI models they can build with no code. Your experts train SharePoint Syntex to understand content like they do, to recognize key information, and to tag content automatically. For example, a contract processing expert can teach SharePoint Syntex to extract the contract’s value, along with the expiration date and key terms and conditions.
This process should result in the information that is provided through SharePoint Syntex being much more relevant across the organization compared to pure machine learning. It could almost be said that this data has a company slant and personality that is provided by the employees building the information set using Power Automate.
SharePoint Syntex will be available as an add-on purchase for commercial Microsoft 365 customers on October 1st, 2020. According to reporting from Mary Jo Foley at ZDNet, that cost is going to be $5 per user per month along with potential volume discounts.
Stay up to speed with all of our Microsoft Ignite 2020 coverage here at ITPro Today for further news and analysis.