Operational Decisions Management 101
-
Upload
alain-neyroud -
Category
Software
-
view
94 -
download
1
Transcript of Operational Decisions Management 101
© 2015 IBM Corporation© 2015 IBM Corporation
Operational Decisions Management 101Alain Neyroud,Program Director, Decision Management IBMPierre Berlandier,STSM, Decision Management IBM
© 2015 IBM Corporation
Decisions 101
• Setting the Stage• Operational Decision Management• Decision Engineering• A few more considerations
© 2015 IBM Corporation
Setting the StageOperational Decision Management 101
3
© 2015 IBM Corporation
What are we discussing
4
Decision A decision is the selection between possible actions.
Decision Making
Decision making can be regarded as the (cognitive) processes resulting in the selection of a course of action among several alternative scenarios. Every decision making process produces a final choice.
Decision Automation Decision automation is considered most appropriate for well-structured, clearly defined, routine or programmed (cf., Simon, 1960) decision situations.
Decision Management Decision management is a business discipline that enables organizations to automate, optimize, and govern repeatable business decisions
© 2015 IBM Corporation
Candidate Decisions
5
Strategic
Tactical
Operational
Frequency of OccurenceLow
Valu
e of
a s
ingl
e de
cisi
on
HighLow
High Candidates for Decision Automation and Management
Operational Decision are repeatable, lower value decisions that typically relate to a single customer or a single transaction. Because of the number of times they must be made, consistency and repeatability are critical.
A repeatable decision is one that is made more than once by an organization following a well-defined, or at least definable, decision-making approach (Taylor, James (2011-10-13)).
© 2015 IBM Corporation
Decision Variations and Change Dynamic
Decision Variations drive the complexity of the decision making process Decision Variations implies related change dynamics
– Few variations => few changes
– Many variations => Many changes
Decision Automation must support decision change dynamics– Otherwise the automated decision quickly become obsolete
Change Dynamics
Market Dynamics
Competition
Merger and acquisition
Regulation
…DecisionVariation
s
Jurisdiction
Product
Channel
Market Segment
…
© 2015 IBM Corporation
When should you consider Operational Decision Management
• Significant Decision Variations
• Significant Change Dynamics
• Other influencing factors– Visibility– Traceability– Auditability
Change Dynamics
Dec
isio
n Va
riatio
n
© 2015 IBM Corporation
Qualifying Changes applied to Operational Decisions
Disruptive
Evolutive
Adaptative
Ex: update existing rate sheet
Ex: provide new typeof vehicle coverage
Ex: enter new LOB(auto vs property)
Impact of change
Freq
uenc
y
Run the business
Change the business
Low HighLow
High
© 2015 IBM Corporation
High
9
Anticipating the nature of Decision Changes
Disruptive
Evolutive
Adaptative
Ex: update existing rate sheet
Ex: provide new typeof vehicle coverage
Ex: enter new LOB(auto vs property)
Operation (80%)
Engineering (20%)
Predictability(ability to anticipate the change in advance)
Freq
uenc
y
Low HighLow
© 2015 IBM Corporation
Take Away
• Operational Decisions are highly repeatable and therefore are good candidates for decision automation
• Decision variations and related change dynamics define the complexity of the decision making process associated with operational decisions
• The ability to manage decision changes is the most important factor to successful operational decision automation
• The majority of decision changes are adaptive in nature:– They are highly repeatable– Their impact is well understood
10
© 2015 IBM Corporation
Operational Decision Management
© 2015 IBM Corporation
Why anticipate the changes?
• Software engineering is about creating application capabilities that support defined Business requirements
• It is unlikely that an application will provide capabilities that support Business requirements that were NOT specified initially
• If Change is not a built-in capability of an application, it has to happen at engineering time
• In order for Change to be a built-in capability of an application, change requirements have to be specified initially
12
© 2015 IBM Corporation 13
Classical IT Change Management
13
Manage and Monitor
InitialRequirements
DesignConstruct
Test
Deploy
Software Development Lifecycle
> 1-3 months
Platform upgrade
DesignConstruct
Test
Deploy
Functional Enhancement
Decision Change
© 2015 IBM Corporation 14
Operational Decision Management
Manage and Monitor
InitialRequirements
DesignConstruct
Test
Deploy
Software Development
> 1-3 months
Platform upgrade
DesignConstruct
Test
Deploy
Functional Enhancement
Business Rule Change
AnalyzeAuthor
Validate
Initial Rules
Deploy
AnalyzeAuthor
Validate
Deploy
AnalyzeAuthor
Validate
Deploy
Operational Decision Management
Change Request
AnalyzeAuthor
Validate
Deploy
< 1 week
Change Request
Change Request
© 2015 IBM Corporation
Traditional Software LifecycleBusiness IT Dev / QAAnalyst Support
Inception Elaboration Construction Transition Production
I1 E1 E2 T1 C2 C3 C1 Application Maintenance
Build-Time Run-Time
© 2015 IBM Corporation
Operational Decision Management Lifecycle
Inception Elaboration Construction Transition Production
I1 E1 T1 C1 Decision Change Management
Build-Time Change-Time
E1 C1 C1
© 2015 IBM Corporation 17
Decision Change Management
Making the change process Easy, Safe and Predictable
Analyze
Author
Validate
Deploy
© 2015 IBM Corporation
The Outcome: Operational change management
• Explicitly support a well defined set of change scenarios (change cases)
• Executing a change cycle is relevant to business engineering,
• Executing a change cycle is NOT relevant to software engineering
• Change cycle shall be executed without interrupting system operation (no delivery of software)
• Change cycle shall preferably be executed by operation personnel (as opposed to engineering personnel) 18
Analyze
Author
Validate
Deploy
Collaboration
Gov
erna
nce
EasySafe
Reliable
© 2015 IBM Corporation
The Outcome: Operational change management• Explicitly support a well defined set of
change scenarios (change cases)• Executing a change cycle is relevant to
business engineering, • Executing a change cycle is NOT
relevant to software engineering• Change cycle shall be executed
without interrupting system operation (no delivery of software)
• Change cycle shall preferably be executed by operation personnel (as opposed to engineering personnel)
19
Analyze
Author
Validate
Deploy
Collaboration
Gov
erna
nce
EasySafe
Predictable
© 2015 IBM Corporation
Take away
• Adaptive changes occur frequently, it is desirable, if not mandatory, for these changes to be supported at operation time
• Adaptive changes and their related impact can be anticipated and it is possible for these changes to be supported at operation time
• Disruptive changes and their associated impact cannot usually be anticipated, it is only possible to support these changes at engineering time
• Operational Change Management provide the framework to manage decision changes at the operation level through a simple, safe and reliable process
© 2015 IBM Corporation
Decision Engineering
© 2015 IBM Corporation
Decision Engineering
• Each engineering cycle goal is to support one or more change case• Every aspects of the change case must be addressed in the cycle (e.g.
analysis, authoring, V&V, deployment, governance)
ProductionInception Elaboration/Construction
Characterize Decision
Evaluate Decision
Variability
DesignRule Model
Define the Verification & Validation
Process
Define the Governance
Process
Define Change Cases
Analyze
Author
Validate
Deploy
Collaboration
Gov
erna
nce
EasySafe
Reliable
Decision CenterBWL Rule Designer
© 2015 IBM Corporation
Decision Discovery: The classic approach
• Find the Business Knowledge• Structure the Knowledge into English Language• Apply analysis to the knowledge to form it into rules• Document the rules, the rule sources, and the vocabulary used for the
rules.
23
© 2015 IBM Corporation 24
Business Rules – a Mean to an End• “Business rules are an expression of business policy in a form that is
understandable to business users and that can be executed by a rule engine. From a business perspective, a business rule is a precise statement that describes, constrains, or controls some aspect of your business. From the IT perspective, business rules are a package of executable business policy statements that can be called from an application.”
03/05/2023 © ILOG, All rights reserved 24
IF the event type equals ‘payment transaction’ and the customer type is ‘corporate’ and the channel used is ‘the web’THEN apply a 30% discount to the standard fee
© 2015 IBM Corporation
Decision Discovery: You don’t need all the rules
• Many decision discovery and analysis methods focus only on the formalization of decisions and rules and promote and exhaustive approach
• However, Decision changes over time so, by the time all rules are discovered, it is likely that a good proportion of them are already outdated
• Moreover, no effort is made to anticipate what type of change will occur in the future
• As a consequence, the resulting application may contain outdated rules and will not likely to be easier to change than any other system
© 2015 IBM Corporation
Change Management Approach to Decision Discovery
One need to discover just enough rules to:• Characterize the decision model functionally
– Ex: what step are used to establish product eligibility
• Determine the variations on this decision model– Ex: product eligibility varies by product, by geography and by channel
• Establish the associated change dynamic– Ex: What trigger change in eligibility along those various dimensions
• Document the anticipated change case– Ex: Add a product, Add a state, update a product, update a state, etc.
• Specify how a change case shall be validated– Ex: establish how a new eligibility guideline should be verified
• Specify how a change case shall be governed– Ex: establish who is participating in a specific change case
© 2015 IBM Corporation
Decision Engineering: The full picture
3 Major phases:• Formalize the decision model and the associated change cases• Design and implement an associated change management framework• Capture and validate the rules in the context of this framework
Inception Elaboration Construction Transition Production
I1 E1 T1 C1 Policy Change Management
Build-Time Change-Time
E1 C1 C1
Implement Decision Framework
Decision Capture and ImplementationFormalize Decision Model and Change Cases
© 2015 IBM Corporation
BRM: Who Can/Shall Participate
Business
IT
Analyze
Author
Validate
Deploy
Analyze
Author
Validate
Deploy
Analyze
Author
Validate
Deploy
Time
Mat
urity
© 2015 IBM Corporation
How do we handle unanticipated change
29
QualificationUnanticipatedChangeRequest
Engineering supervised/executed Adhoc Change
Engineering of new Change Case
Will Repeat
One shot
If it is anticipated that this change will re-occur in the future then a decision engineering cycle should be executed to implement the new change case. Once the change case is implemented, change can occur at decision management time.
If it cannot be anticipated that this change will re-occur in the future then the change shall be implemented within an engineering cycle
© 2015 IBM Corporation
Take away
• Executing a Decision Engineering project is not (just) about discovering and implementing all the rules at engineering time
• Executing a Decision Engineering project is about building a Decision Framework that will enable subject matter expert to discover, capture and change their decisions with minimum support from IT