People do things in a professional organization because it’s their job to do so. Modeling this fundamental fact of organizational life is how an Engaging OS system is animated.

Both people and AI agents access the system as users. In most systems, a user gains privileges by being directly assigned to a role, which behaves like a group. In Engaging OS however, users are not given direct access to the items they work on; rather, types of users are given types of access to types of items.

Access is via positions, such as Financial Manager or Registrar, Ambassador or Electrician. A position is instantiated whenever a user, like an adventurer donning a costume, assumes a role within an organizational body.

A user can hold multiple roles; a position can inherit from other positions; and bodies can encompass other bodies. Such expressiveness enables efficient organizational modeling and configuration, especially at scale. 

A position provides access to screens and operations. An operation consists, at minimum, of three elements:

• an action (e.g., “approve”) 
• a collection (e.g., “invoices”) 
• a state (e.g., “submitted”)

Actions are paired with states to form mutations—so to “approve” an “invoice” always sets its state to “approved”. No change occurs to system data unless resulting from an instance of an operation. And coming soon to enrich mutations are attestations, requiring more than one user to sign off before the mutation can occur.

Operations can also be conditioned on factors beyond state, including the values of other fields in the current item, of environmental values, and—like flicking on a warp drive—values from other items.

AI can therefore be applied safely throughout the system, its inhuman intelligence, scale and speed harnessed to assigned roles, each having explicit sets of capabilities.