|  | 			WHAT IS THE EDGE PROJECT?
EDGE Stand For:  Event Driven Goal Environment
About EDGE:
   - A focused, insurance project (with a customer) which evolves
     into a set of software products
   - Using partnership and (rapid) prototyping to develop product
     requirements
   - Focuses on largely-unautomated underwriting function, with
     capability to expand into (1) agency/branch and (2) policy
     service
   - Achieves first ownership of data, gradually passing less and
     less to mainframe
   - A work-in-process architecture for distributed data & function
   - Combines conventional and expert systems technology in a more
     cogent environment
Why Do This:
   - Deliver superior customer solutions & satisfaction
   - Develop a superior and extensible solutions architecture
   - Capitalize on insurance interest in expert systems
   - Take the first steps to wrest desktop from IBM
   - Position ourselves to move into mainframe territory
   More To Follow
    John 
 | 
|  | 		EDGE FUNCTIONAL COMPONENTS (Pre-Phase Review)
1. W.I.P.S. (Work-In-Process-System) CONTROLLER:  A generic and extensible
   architecture for goal-oriented, event/task processing within a logical
   system which can span multiple, physical, networked systems.
	a. Goal Definition File
	b. Goal Interpreter
	c. Resource & Event Queues
	d. Communications Control Interface Library
	e. Event Processor 
2. RELATIONAL UNDERWRITING DATA BASE:  An RDB repository for all codified
   and textual information captured from (1) applications for insurance, 
   (2) existing internal records and (3) outside information resulting from 
   inquiries, tests and investigatons.  There must be a separate RDB design
   for each line of insurance business, the first of which is individual 
   life.  Major data relationships involve linking the following catagories
   of information.
	a. Case Information
	b. Application Information
	d. Producer Information
	e. Underwriting Information
	f. Host Records
	g. Insurance Plan Information
	h. WIP Status Information
3. FUNCTIONAL SOFTWARE AGENTS:  Done by line of insurance business (the 
   first being individual life), these consists of highly-granular, plug-
   in/out "modules" for processing an application (i.e. app) from point
   of sale to issue of policy.  The following software agents apply to 
   the Phoenix prototype and are a representative subset of those that
   apply to a future (individual life) product specification or set.  
	app data entry          app maintenance         app validation
	plan validation         producer validation     app review
	requirements gen        mark as validated       submit app
	order requirements      order APS               order lab
	order Equifax           APS requirements        info needs notify
	field office menu       resolve requirements    reverse status
	mail notification       reports                 help
	remote data x-fer       message receive         sig events entry
	sig events inquiry      decnet startup          HO DB merge
	requirements orderer    alpha search            alpha update
	ralph                   APS swa                 NBI order
	resolve requirements    follow-up               assign pol. #
	assign underwriter      case presentation       und. APS decision
	underwriting decision   pre-underwr. filter     user req. order
	post approval           issue requirements      delivery req'ments
	issue sys feed          issue data receive      update case as paid
	feed instat             MIB request             MIB reply
   NOTE:
	1. More detailed descriptions of the functionality of these is both 
	   available and undergoing further development.  The list is yet
	   incomplete.
	2. Many of the above software agents relate to post-sale, pre-
	   underwriting work done in the agency/branch office and will
	   process in those physical systems.
	3. Several of the above use expert systems.
	4. Not all of the above will be completed for the prototype.
John McAtee
 |