Microsoft has positioned the Microsoft Graph and its 467M users of the LinkedIn Graph as the “most strategic data asset” which will enable IT departments to deliver AI to information workers. Box has also announced its own enterprise graph to enable intelligent content recommendations for business users. With the announcement of how graphs will enable enterprise AI, it is time to take a deeper look at its business benefits.
What is an enterprise graph?
Since most enterprise users connect to the cloud to perform their daily business tasks using services such as Office 365, Box or Salesforce, these business cloud vendors can, for the first time, build a dynamic map of relationships between people and cloud events. To understand who is working together, traditional software depends on data entry from users, like by building list of favorites or email distribution lists. But how often do you update your favorites on your phone or on your messaging app? Intelligent cloud software senses these events and continuously updates the evolving dependency graph between people and content.
The enterprise graph includes three types of nodes: people, topics and cloud events.
Cloud events are anything happening in the cloud. For example, a new opportunity created in the CRM system, an update of a presentation you have been contributing to or a new video which has been created by your marketing staff. Every time you connect with a colleague, upload a new document or update a sales opportunity, your enterprise graph is being updated.
Topics are the next level of abstraction above cloud events. They are typically proper nouns, like a project name, a campaign name or a key customer name. Most cloud events are connected to one or more topics within your organization. Topics enable the grouping together of related cloud events, unveiling the dependency of events across cloud services.
In the example above, you can see how people in the organizations are collaborating on three Topics: Product Launch, Launch Event and Reference Customer, and the cloud events which are associated to these three topics.
As the graph dynamically learns and updates the dependencies between people, topics and cloud events, it enables the delivery of smart notifications. In our example, once the creative lead has updated the product screenshots before the product launch, the account manager gets notified that he might want to update his proposal with the latest screenshots, which reflect the changes requested by the reference customer. Since the proposal and the screenshots reside in different clouds, traditional cloud services would not capture the dependency between the two. Only a cross-cloud, cross-vendor enterprise graph can capture the dependency between the proposal and the screenshots, as they are connected through the same topic, the “reference customer,” and then deliver meaningful notifications. Smart notifications reduce the cognitive burden on information workers since they only get insightful and timely notifications instead of the current overwhelming app notification tsunami.
Focus on what matters most
An enterprise graph can also help information workers focus. Rather than promoting distractions by delivering a stream of cloud updates organized by apps, an enterprise graph delivers information by topics. For the marketing manager, the stream of events related to the “product launch” topic will include the four adjacent nodes to the “product launch” node. Specifically, the new email template, the updated launch pitch presentation, the draft email for the product announcement and the screenshots that the documentation lead just updated.
An enterprise graph that incorporates the relationships between people, topics and cloud events provides the foundation of the next generation of cross-cloud, intelligent applications. This enables information workers to discover content the way the human brain works – by topics – rather than each app being an additional data silo with a limited view on a given topic.
This article was originally published on ComputerWorld.