1 Decision Automation using PowerPoint Presentation, PPT - DocSlides

1

Decision Automation

using Models , Services and Dashboards

Chair BPMN MIWG at OMGCo-Chair BPM Health Workgroup at OMGDMN RTF Member at OMGBPMN RTF Member at OMGCMMN RTF Member at OMGChair BPSWG at WfMCXPDL Co-Editor at WfMC

Denis GagnéCEO & CTO

Abstract

The

Decision Model and Notation (DMN) offers the perfect solution to specifying Business Decisions. Symbolic and sub-symbolic artificial Intelligence (AI) approaches can effectively be used jointly within DMN, delivering explainable decision automation that is desired if not mandatory in any business context. The resulting DMN Decision Models are also the perfect architecture for the creation of decision support dashboards.

In this session we will demonstrate how line of business people can define business decisions that are explainable, auditable, and traceable. These business decisions can be assembled and consumed as services via a modern platform API architecture and visualized via graphical dashboards. The resulting dashboards help visualize information and knowledge that that are critical for the operations of any type of business.

Digital businesses are insights driven

To turn insights into actions, they rely on their ability to make the right business decisions

🐘

Analytics and Decisions

4

Gartner 2013

5

A decision that's made by technology that might have been made by a human

Decision automationSimplicable.org

🐘

Technology = Hardcoded rules or complex algorithm

AI

6

🐘

To have computers do things, that, if people did them, we would consider intelligentTo explain how human intelligence works, and reproduce it in computersGoals of AIsubject to “Disappearing AI”👆

AI the elephant in the room

The Dark Secret

No one really knows how the most advanced algorithms do what they do.

That could be a problem.at the Heart of AI

7MIT Technology Review 2017

💣

🐘

8

Place your screenshot here

Marvin Minsky

AI

The current AI

resurgence is

connectionist

(sub symbolic)

driven

Explainable

AI is required if not mandatoryA Black Box solution will not cut it

9

Business

Decisions

🐘

Explainable

Decisions

Transparency, Provability and

Explainabilityare mandatory requirements of today’s business decisions automationArticle 22 of the GDPR requires that companies be able to explain to their customers how machine decisions are made.

10

11

DMN to the rescue

12

What is DMN?

Decision Model and Notation

DMN is a standard published by the Object Management Group (OMG). Why DMN matters?A common meta-model and notation for describing and modeling repeatable Business Decisions Enables various groups to effectively collaborate in defining a Decision ModelProvides a standard notation for Decision TablesCurrent StatusDMN 1.2 is the current version

https://www.omg.org/dmnDMN

DMN

Decisions

Decision Support

Decision Automation13DMN DMN 👤User acts based on the recommendation from the DMN decisionSystem acts based onthe DMN decision

Decision Automation based on

DMN14

Decisions as first class citizen

Visual representation of decisionsWhat you model is what you execute!Completely explainable decision automation

DMN

is an open StandardDMN is Model Driven

15

Decisions

Tactical

Operational

Strategic16

DMN

IEEE and ISO are exploring the usage of DMN for strategic decision in the context of governance and ethics

What you model is what you execute !

17

The Model is the Decision Requirements

The Model is the Decision LogicThe Model is the Decision DocumentationThe Model is the Decision CodeThe Model is VerifiableDMN

DMNDMN

A picture is worth a thousand words

A complex business decision can be conveyed with a simple Decision Requirement Diagram (DRD), providing context and making the decision and its context obvious and transparent

18

19

DMN Model

Offers an explainable business context to Black Box AI

Identifies Data and Information requirements of Black Box AISpecifies Black Box AI relation to Business Decisions

Frames the impact analysis of Black Box AI on actual Business Decisions 20

auditableexplainabletraceable

🐘

21

Black Box AI

DMN provides business context to black box AI

Dec

ision

Most powerful concept of DMN Enables Decision as a Service (

DaaS)

Services

22

Decision Services

Allows DMN decisions to be easily encapsulated, tested, re-used, and orchestrated in various apps

23

DeterministicStatelessReusableAutonomous units of business functionalityContract-basedLoosely coupled

Platform-independentDiscoverable and location independentStandards-basedAPI

Decision

As a

Service Full lifecycle management platform of Decision Service APIs for composite apps

24

🏭

DMN

API

🏭

Internal Library of services

25

External Store of services

Decision as a Service

26

🏭

DMN

API

Dashboarding

Decision Requirement Diagrams as Dashboard specifications

27

Frame the impact analysis of Information and knowledge requirements on actual Business Decisions

28

DMN

Decisions

DescriptiveWhat Happened?DMN Decision Services execution logs can be configured for audits of WHY and HOW any Business Decision was achieved

DiagnosticWhy did it happen?DMN Decision Services exposed as a Dashboard can provide insight on the Business Decision behavior wrt dependency variationsPrescriptiveWhat should I do?DMN Decision Services can automate Business Decisions that are explainable, auditable and traceable 29

30

31

32

Decision Automation

Decision Models

Decision Services

Decision Dashboards33

DMN

DMN

DMN

DMN

auditable

explainable

traceable

THANKS!

Any questions?You can find me at

dgagne@Trisotech.com

34👦