Microsoft’s Project Cortex Wants You to Hire a Knowledge Manager
With all the talk about AI replacing people in business, here is (perhaps) welcome news. As part of Microsoft’s Project Cortex rollout, Microsoft is advocating for new knowledge management positions, staffed by living, breathing human beings:
Here is a brief description of each role:
Content Services Admin (aka "Knowledge Admin") is an IT role whose responsibilities include setting up the knowledge product suite (i.e. Cortex) and keeping it running as a knowledge platform. This person is also responsible for ensuring security and compliance standards.
Knowledge Manager (KM) is responsible for the quality of knowledge shared across the organization. KMs are responsible for aggregating the taxonomy of the organization by creating, editing, deleting and rejecting topic suggestions provided by subject matter experts. While this is not a new job definition per se, to date, it has been instituted in only a very small number of organizations that have invested heavily in the classification of their assets, like documents and emails. Cortex is proposing that each organization retain a senior-level KM.
Subject Matter Experts are business workers who have a deep understanding of how information is organized in their department or business function. Experts have always existed, but they are now being tasked with organizing and curating knowledge by defining topics for their departments.
In many cases, the Content Services Admin is simply an expanded role for the SharePoint and/or Teams administrator. But the Knowledge Manager is a brand-new position. Most companies do not have a centralized function for compiling knowledge. Cortex, by bringing Microsoft’s clout to the realm of knowledge management, represents a landmark shift in how IT infrastructure will support the business going forward.
Related Article: Microsoft Project Cortex Ushers in the Age of Topic Computing
What’s Different This Time?
Knowledge, for our purposes, refers to the compilation of related information in meaningful ways so that personnel can analyze a business situation and take well-informed action. Presently, it is impractical to aggregate disconnected business information contained in email, documents, events, tasks and contacts because they reside in disparate repositories. And even if it were possible to put everything into a single repository, normalizing the data so they can be analyzed is too complex an operation for most organizations.
Now, with the Microsoft 365 cloud suite of products, emails, documents, events, tasks and contacts are, for the first time, stored in a common data structure and organized meaningfully using the Microsoft Graph. The Graph maintains relationships between content and people by tracking who created, edited and shared each piece of content.
What is currently missing is a contextual relationship between different items in the Graph. Context supplies the business meaning of documents and other items. In Cortex, context is specified through a topic. For example, one of the topics of a project plan document would be the name of the project. Topics are defined by the subject matter experts but rank and file workers need to manually apply topics to individual pieces of content when they create them. This is similar in concept to how workers apply SharePoint metadata to classify documents when they are created. In fact, much of the underlying Cortex topics infrastructure relies upon enhanced SharePoint metadata.
Once topics have been applied to enough documents and other content, AI, machine learning and NLP can be used to automatically apply topics to new content. The result is that, over time information from many applications can be aggregated and grouped by common topics with relatively little human interaction.
The heavy lifting comes from accurately defining the organization’s topics across many departments, keeping the topics current, and ensuring that topics are defined and applied uniformly by a diverse group of workers; workers possessing a range of skills and business acumen. This where the new Knowledge Manager comes in.
Related Article: The Secret Sauce Behind Project Cortex: Good Metadata
The Knowledge Manager, Reimagined in Cortex's Shadow
In the past, the role of knowledge management has been largely associated with information and library sciences. For those few organizations who have retained a knowledge manager, their role has been to help store and classify information so it can be found, rather than to organize knowledge to improve strategic planning and operation.
Now, with Cortex providing organizations with a practical pathway for turning information into knowledge, will businesses hire and empower a senior-level Knowledge Manager so the company can move faster, make better decisions and be more competitive? I believe the answer is, eventually yes. But getting Cortex to work will take several years, because the underlying machinery is complex and getting all the pieces to mesh smoothly will take heavy lifting on Microsoft’s part. It will also require organizations to invest heavily to get their IT infrastructures and workforce Cortex-ready.
Another question is whether the new Knowledge Manager can secure the political clout necessary to define what knowledge means for their entire organization. Political, because whoever controls knowledge, controls the organization. Going forward, it is knowledge, not information, that will impact how budgets are allocated, which products and services are offered, and even who gets promoted.
As such, the new Knowledge Manager must be a senior position to drive changes across many departments. The addition of a new senior information-related position has precedents. The roles of CIO, CSO and CISO haven’t been around for very long, and they already command power in today’s information-driven economy.
Related Article: The State of Knowledge Management in 2020
From an Information Enterprise to a Knowledge Enterprise? Not so Fast
It currently remains to be seen if organizations can make the transformation from "information-driven" to "knowledge-driven" enterprises. And as usual, it is not only about the technology. The transformation will require fundamental changes in employee behavior and business processes. Here are some of basic questions that still need to be answered:
- Will subject matter experts dedicate the time and effort needed to build meaningful topics?
- Will organizations be able to agree on the list of topics that define their business? Defining topics is potentially a political powder keg. Case in point: the CIO of a large federal agency told me a few years ago that after several years of effort to define their organization’s topics, senior management could only agree upon four topics: one of them was the names of the departments, and another was the date the content was created.
- Where does the Knowledge Manager fit in the hierarchy of the CIO’s office? Will they be able to secure the leverage necessary to effect broad changes?
- Will rank and file employees spend the time and effort to associate documents, emails and events with topics defined by the experts? Experience from many years of trying to convince SharePoint users to apply metadata to documents says this will be a monumental task.
- Even if employees do invest the effort, will they be able to apply topics uniformly enough to create a credible learning set for the Cortex AI engine to generate accurate topic suggestions?
- Will the topics defined by the subject matter experts and knowledge manager be intuitive to rank and file workers? Will they be intuitive enough to help workers find related information, see the bigger picture and make better business decisions?
These are only a few of the hurdles ahead. There will be many more. But Rome wasn’t built in a day and the knowledge enterprise won’t be either. We have to start somewhere, and with Microsoft’s control of the enterprise IT infrastructure, Cortex has the best chance for success. So, stay tuned as we start to see early adopter deployments later this year.
About the Author
David is a product expert with extensive experience leading information-intensive technology organizations. His specialty is helping organizations “do it right the first time”— get to market quickly and successfully through a structured process of working closely with design partners from day one.