Successfully Implementing Knowledge-Based IT

Successfully Implementing Knowledge-based IT

Integrating BPM and Rules-Based Technologies

© 2016 Vizuri – All Rights Reserved

Joe Dickman – SVP, Vizuri

Business leader working with organizations to understand how to leverage innovation for increased business agility. Specializes in insurance, healthcare, financial service, retail, and logistics.

Ken Spokas – Technical Director, VizuriTechnology and Subject Matter Expert working with organizations to evaluate architectural approaches, design extensible solutions, and introduce innovative technologies to improve business agility.

Introductions


What will we learn today?

• How efficiencies can be gained throughthe alignment and integration of Business Process Management (BPM) with Business Rules Management Systems (BRMS)

• Discuss practical and field-proven approaches that take a more holistic approach at defining business processes that affect business across the the enterprise


Building a Bridge between Goals and Outcomes

Building this bridge has some challenges

• The challenges we expect to encounter by implementing BPM and Rules-Based systems are:

– Shared Vocabulary– Formally defined rules and processes– Business and IT Champions– 360 degree feedback loop– Organizational visibility across the entire business


What the business envisioned

…and what IT delivered...

Expectations vs. Outcomes

Business groups and IT require new positions to be defined that ”bridge” the communication gap.


What is Knowledge-based IT?A knowledge-based system (KBS) focuses on systems that use knowledge-based techniques to support human decision-making, learning and action1

Knowledge-based IT (KB-IT) is the application of KBS techniques with information technology• Design process• Models and methods• Software tools• Decision-support mechanisms • Digital representation, and • System architectures

Source: Fujita, H., & Lu, J. (2012). Knowledge-based Systems, Volume(31), pp. 1-211.


What is Knowledge-based IT?

IT Solution where Knowledge is a first class citizen– Business Rules and Process boundaries are clearly defined– Goals drive the solution, not the IT restrictions

Change can be accommodated quickly without being impeded by IT limitations

– The process knowledge definition and concept model should ”outlive” any technical implementation

– Decouple the concept model and process definition from the underlying implementation to create business agility


BPM, business processes and rules

• BPM provides a systematic approach to making an organization's workflow more effective, efficient, and capable of adapting to ever-changing business conditions

• A business process is an activity or set of activities that will accomplish a specific organizational goal

• A business rule management system (BRMS) enables organizational policies, and the operational decisions associated with those policies, to be defined, deployed, monitored and maintained separately from core application code


Value of BPM and Rules Technology

• Unlock Rules from underlying technology

• Enforce business processes and rules consistently

• Ensure compliance to organizational and regulatory requirements

• Gain visibility into business operations• Enable decision management in real-time through algorithmic

approaches• Eliminate the business risks associated with “Adverse

Selection”


When should you use a Rules Engine?

Complex Problems• Highly conditional• Logic prone to change• High-level of subject matter

expertise required• Goal is new, still being defined

Traditional code approach not working well• Problem is too fragile to approach

procedurally• Analyst needs insight into implementation


What is your definition of BPM Automation?

BPM Automation has many different meanings based on stakeholders viewpoint. It can be:• an Oversight engine

loosely defines and manages a business process,

• an Orchestration engine manages low-level interactions betweensystems, or

• a Platformmanage the lifecycle of documents.

BPM Suite


When should you use BPM?

BPM Suite

AHA!


The Nature of BPM Automation

When considering the implementation of BPM Automation keep in mind that it’s a “declarative” approach

• Processes represent the definition, structure, and flow of linked tasks, but should not include any implementation

• Majority of task work should be conducted outside of the process definition and invoked (e.g. asynchronous messaging, web services, etc.)

• Processes spend the overwhelming majority of their time “sleeping” while waiting on a signal or human task to complete


When should you use BPM?

A primary goal is often to define a unified “body of knowledge” across the organization• Increase organizational agility• Document processes to identify gaps

and critical dependencies• Protect Intellectual Capital• Reduce Complexity• Provide insight into implementation• Improve IT Outcomes• Collect Key Performance Indicators• Support Contingency Planning


Compliance, Audit, and Policy Related Business Processes• Insurance

– Underwriting– Claim Adjudication

• Healthcare– Standards of Care– Plans of Action

• HIPAA• SEC Finance Disclosure• Mortgage Lending• Government Agencies

Related Business Case Examples

Dynamic Insurance Questionnaire ExampleUse Case


BPM tools manage the Business Process and BRMS tools focus on Liability Management

• Identification and application of claims rules improvesactuarial tables and underwritingaccuracy

• BRMS technology supports the identification of high-risk groups

• Mitigate adverse selection duringunderwriting and fraud detection in claim submission processes

Insurance Rules-Based Dynamic Questionnaire


Dynamic Questionnaire – Single Page Web Application


Dynamic Questionnaire Design

Question Definitions

Answers that trigger question

state

Validation Feedback

Rate Calculation

The design utilizes a BPM Process workflow dynamically driven by a Rules domain


Questionnaire Requirements

Questionnaire represents a Process that utilizes a Rules Domainto present, collect, and evaluate responses• Questionnaire Definition

– Powers the view (order, type, required, validation, etc.– Questions can be added/updated without coding

• Dynamic Question State– Answers may hide/show/change questions

• Validation Feedback– Simple validation (e.g. correct data type, required values, etc.)– “Do Not Bind” decision points (i.e. State not covered)– Risk categorization

• Rate Calculation– Based on information collected, risks and attributes, what is the “Policy Rate

Standards become critical enablers at the data, process, and infrastructure level


Organizing Rule Sets

Identified Phases that Rules should be grouped in1. Questionnaire Construction – Define “blank” questionnaire2. Question Display – Changes to questions based on answers3. Feedback – Validation messages/errors for answers4. Rating Calculation – Policy cost determination

• Categories can be co-located at design time• Can execute in isolation at run time• Share a common domain model…


Domain Model – Nouns of the Rules

Question• Generic Question Holder• Describes display order, data type, is required, etc.• Hierarchical: “sub-questions”

Property and Address• Holders for answers with highly specific attributes

(e.g. roof age, has swimming pool, zip code, etc.)• Allows expressive rule authoring for display phase


Domain Model – An Interlude

“Generic” Answer Model• When $answer : Answer (questionid == 15, interger > 25)• Then…

“Specific” Answer Model• When $property: Property (roofAge >25) • Then…

© 2016 Vizuri – All Rights Reserved© 2015 Vizuri – All Rights Reserved

Quote Message• Generic holder for feedback on answers• Different severities: INFO, WARNING, ERROR• Optionally tied to a specific data point

Quote• Simple holder for calculated premium

Domain Model – Nouns of the Rules continued


Anatomy of an actual rule1. Attributes – names, properties, descriptions, etc.

2. When – Condition that activates rule

3. Then – Consequence occurs when activated

• Basically, a disconnected ”if then” statement• Declarative nature promotes breaking a problem down• Singularly focused, powerful as a group

Rules Authoring – Technical Rules

Implementation Best Practice Recommendations


Not tied to a specific Technology Solution

But a specific “type” of technology solution– Business Process Management (BPM) Products– Business Rules Management Systems (BRMS)

Products

Tools must support the ability to be Declarative in nature, not the typical Imperative coding styles


Best Practices From the Field

Get Organized All Around!Both IT and Business Need Coordinated Discipline

1. Take an inventory of the AS-IS problems and solutions2. Get Commitment!3. Build a common way to describe things4. Decompose the problem appropriately5. Build with quick iterations6. Continually monitor, adapt and improve

A little discipline at each step pays off in both the short and long term!



Step 1: Take an Inventory – Business SideDoes the problem lend itself to using Knowledge Based IT by exhibiting inherent complexity of the business logic:

– Problem is by nature highly conditional (e.g., many permutations of inputs).

– Problem requires a high-level of Subject Matter Expertise (SME).

– “Maturity” of problem definition (i.e., is this still being figured out).



Step 1: Take an Inventory – Business SideDoes the problem lend itself to using Knowledge Based IT by requiring visibility:

– Need for insight into execution (e.g., why did we get X answer with Y inputs).

– Need for all interested stakeholders to understand implementation.



Step 1: Take an Inventory – Business SideDoes the problem lend itself to using Knowledge Based IT by requiring consistency:

– Need for problem to be well understood by all stakeholders.

– Concept Model exists and is agreed upon by all stakeholders.



Step 1: Take an Inventory – Business SideDoes the problem lend itself to using Knowledge Based IT by requiring agility:

– Business logic changes unpredictably (i.e., “off-cycle” changes required).

– Business logic changes need to be made quickly.– Current implementation is “fragile” (i.e., changes cannot

be made with confidence).



Step 1: Take an Inventory – Business SideMake a report card on how applicable each factor is:



Step 1: Take an Inventory – IT SideAssess the implementation on the following high-level criteria:

• The critical business logic is consistent throughout its lifecycle within the project.

– Minimal “translation” layers as it is harvested, implemented and tested.• This lifecycle provides an acceptable degree of agility for the project’s

release cycle.– Change is quick to introduce and doesn’t require restructuring of code.

• The implementation of the logic is visible to those who are responsible for it.• Pure form in a clear and concise manner to non/semi-technical resources



Step 1: Take an Inventory – IT SidePrioritize improvements on a combination of severity and effort



Step 2: Get CommitmentNeed all stakeholders to make a sincere commitment to improve:

• The more management buy-in, the more hope of project success– Time-boxed modernization effort (typically 3 to 5 years)– Appetite to continue support of center of excellence beyond inception

• Need influential champions on both Business and IT side– IT left alone will mangle Knowledge IT to look like traditional IT– Business left alone will make sweeping changes without bound



Step 3: Get SynchronizedAll stakeholders must be able to communicate effectively:

• Common Vocabulary– Non-ambiguous, understood by all– Understood, or in a central reference– Composes the “Concept Model”

• Business Rules succinctly defined in near human language– Not necessarily executable– Closest to real language the better, but avoiding ambiguity

• High level “happy path” processes modeled visually



Step 4: Plan IT ImplementationBuild systems that effectively separate the plumbing from Knowledge:

• Select frameworks that support declarative logic and processes– Business Process Management (BPM) platforms that can build

upon formal process definitions– Business Rules Management Systems (BRMS) that can model and

execute logic in near natural language formats• Make sure they support multiple concurrent versions of

decisions/processes• Consider small, precise tools rather than all-in-one solutions

– Wrap interactions in business-centric APIs



Step 4: Plan IT ImplementationDecompose the solution in appropriate layers:

• BPM projects for instance, are best conceptualized in a tiered approach:

– Idealistic (non-executable) Process Definition– Complete (executable) Process Implementation– Enterprise Integration Management (EIM) Layer



Step 4: Plan IT ImplementationDecompose the components of the solution in appropriate size:

• Current buzzword is “microservices”• Making a concerted effort to avoid monolithic solutions• Sized to swap out components without disrupting others• Can be harder to get this right than you may think!



Step 5: Iterate QuicklyBuild as quickly as possible, with many feedback loops:

• Choose agile software development methodology– Time boxes iterations to several weeks– Show and tell at the end of every iteration

• Create throw-away user interfaces that illustrate the intent• These spark buy-in and feedback• Often become part of the more permanent solution• Achieved through exposing the complexity through a business

centric interface that can be mocked before fully implemented



Step 6: Monitor, Adapt and ImproveBuild the system so that it exposes the important details:

• Processes must be defined in terms of Key Performance Indicators (KPIs)

– Select frameworks that expose these metrics– Make dashboards so they can be continually monitored

• Rules should be auditable from a business perspective– Must explain why a particular decision was made– Simulation environment comparing results of 2 different versions

of the rule set is invaluable in enabling business to have the confidence to continually improve the rules


Summary

Build the system for visibility, consistency and agility at each stage:

– Design

• common vocabulary and formal definition

– Implementation

• understandable from non-technical individuals

• Enables ability to change quickly

– Execution

• can determine how an outcome was determined

– Improvement

• necessary metrics are exposed to drive change


© 2016 Vizuri

Joe DickmanVizuri

[email protected]

Ken SpokasVizuri

[email protected]

Thank You

Visit us on the Exhibit Floor @ the Red Hat Booth

Successfully Implementing Knowledge-Based IT

Software

Transcript of Successfully Implementing Knowledge-Based IT