Assurance Leadership Forum (ALF) - Cognitive Adaptive Testing
-
Upload
jonathon-wright -
Category
Technology
-
view
676 -
download
0
Transcript of Assurance Leadership Forum (ALF) - Cognitive Adaptive Testing
1
Cognitive Adaptive Testing based on analytics-driven digital experiences
Jonathon WrightAutomation, DevOps & Digital [email protected] | linkedin.com/in/automation
2
Abstract
Jonathon Wright
DevOps, Automation & Digital Assurance
3
The success of our technologies largely depend
on the fact that, while we were speculating about the
possibility of ultra-intelligence, we increasingly
enveloped the world in so many devices, sensors,
application and data that it became an IT-friendly
environment, where technologies could replace us
without having any understanding, intentions,
interpretations, emotional states, semantic skills,
consciousness, self-awareness or flexible intelligence.
Memory as in algorithms and immense datasets now
outperforms intelligence. MegaTech, Technology in 2050, The Economist, Daniel Franklin, 2015
“
”
THE NEED FOR SPEED
94%of executives face increased pressure to release appsmore quickly
“
Journey to the Digital Testing
3
Based on 2014 Vanson Bourne study commissioned by CA
4
Digital Testing, Quality & Assurance
DIGITAL TESTING, QUALITY & ASSURANCE
http://bit.ly/digital-assurance http://a.co/jlxMHWp
5
DIGITAL TESTING, QUALITY & ASSURANCE
Digital Testing, Quality & Assurance
http://www.docjamesw.com https://youtu.be/e3PzRnY-zTg
6
‘Shift Right’ – Performance Lifecycle
7
IdeasBeliefs, Values, Culture
Heuristics
Behaviors
Value
AgileLean
Continuous Testing 101Webcast Series – Value-Driven Delivery – Episode I
Methods and Skills
http://catechnologies.podbean.comRay Arell
8
Shift Left - ResearchModel Driven Delivery (MDD)
SHIFT LEFT
https://www.ca.com/content/dam/ca/us/files/industry-analyst-report/continuous-testing-as-a-digital-business-enabler.pdf
Number of custom AUT
9
Digital TestingUrban 4.0 - Predictive Crime feat. video analytics
SHIFT RIGHT
https://www.slideshare.net/Jonathon_Wright/oracle-digital-evolution-enterprise-of-things
10
Cognitive Adaptive Testing (CAT)
11
Digital Testing
TESTING IN THE DIGITAL WORLD
COGNITIVE ADAPTIVE ADOPTION (DESIGNOPS)
COGNITIVE ADAPTIVE DELIVERY (VALUE-DRIVEN)
COGNITIVE ADAPTIVE LEARNING (ANALYTICS-DRIVEN)
COGNITIVE ADAPTIVE TECHNOLOGY (INSIGHT-DRIVEN)
WHERE NEXT?
1
2
3
4
5
6
12 © 2016 CA. ALL RIGHTS RESERVED.12
Portfolio Management
Unify long term strategy, investment and portfolio planning.
Collaboratively plan, prioritize and track work across the enterprise.
Project ManagementDevelop & Test
Synchronize efforts to dramatically speed app development & increase quality
Manage & Monitor
Make a great customer experience a competitive advantage
Release & Deploy
Control the release process, to continuously advance application quality, improve the customer experience and reduce costs.
ContinuousDelivery
Minimum Viable Experience
Minimum Viable Quality
Value-Driven Delivery
BUSINESS INITIATIVE
Complexity-Informed Organizational Change
BUSINESS VALUE
MinimumViableProduct
CUSTOMER
THINK
MAKE
Digital TestingComplexity-Informed Organizational Change
13
DesignOpsCreate and deliver Ideas to Outcomes
Pervasive Security
Redefine how work is planned, managed and
executed
Agile Management
Provide a flawless app experience optimized for
performance
Agile Operations
Accelerate and streamline development, testing and
release
Continuous Delivery
Continuous Development Continuous Adaptive Testing Continuous Release
Digital TestingValue-Driven Delivery (Shift-Left & Shift-Right)
14
DIGITAL ASSURANCE & TESTING
VALUE-DRIVEN DELIVERY
1Cognitive Adaptive Delivery
2Cognitive Adaptive Learning
4Cognitive Adaptive
Technology
3Cognitive AdaptiveTesting
Digital TestingCognitive Adaptive Adoption (CA-A)
16
The ability to reliably release high-quality solutions at any time
COGNITIVE ADAPTIVE TESTING
Accelerate Deliveries
Reduce Errors
Manage Complexity
Increase Visibility
Drive Collaboration
Continually Improve
Digital TestingContinuous Adaptive Testing (CAT)
17
> 50%large organizations globally will compete using advanced analytics & proprietary algorithms
40%enterprises' net-new analytics investment will go to predictive / prescriptive
75%technology-oriented business intelligence competency centers will evolve to focus on information value generation
Source: Gartner
Digital TestingIntelligent Analytics is the New Battleground
18
SOI
Visualization
(Big) Data
AnalyticsDriven
Inference (Machine
Learning & AI)
Intelligent Process
Automation
Digital Technology Integrations
Digital TestingSystems of Intelligence
18
Inspired by “Intelligent DevOps – Driving Continuous Delivery by Harnessing the Power of Analytics & Autonomics”, Shamim Ahmed, 11/16/2016, CA World 2016, Las Vegas
19
Digital TestingWhat is analytics-driven quality?
Model ProgressiveRefinement
Analytics-Driven
ActionsPredictionsMeta Data
Inspired by “Intelligent DevOps – Driving Continuous Delivery by Harnessing the Power of Analytics & Autonomics”, Shamim Ahmed, 11/16/2016, CA World 2016, Las Vegas
20
Portfolio backlog prioritization and
value analysis
Improved scheduling and
capacity
Validated requirements in
backlog
Improved estimates
Predict failures based on code
commits
Dynamic code promotion
criteria
Predict defects/failures
Optimized Tests
Scenario identification
based on trends
DX Scores
Anomaly detection
PLAN DEFINE DEVELOP TEST RELEASE OPERATEPhase
Examples of Cognitive Adaptive
Intelligence
Examples of Outcomes
Investment decision analytics
Code quality
analytics
TestOptimization
& Defectprediction
A/B Testing
& Canary
Log & DX Analytics
DX-basedrequirements
Inspired by “Intelligent DevOps – Driving Continuous Delivery by Harnessing the Power of Analytics & Autonomics”, Shamim Ahmed, 11/16/2016, CA World 2016, Las Vegas
Digital Testing
Cognitive Adaptive Intelligence (CAi)
Systems of Intelligence
21
TEST MAINTENANCE STATIC RULES BIG (un-structured) DATA
IOT, IOE
CX / DX SECURITY
STRUCTUREDAUTOMATION
Digital TestingKey Challenges Posed by Adaptive Digitalization
Inspired by “Intelligent DevOps – Driving Continuous Delivery by Harnessing the Power of Analytics & Autonomics”, Shamim Ahmed, 11/16/2016, CA World 2016, Las Vegas
22
▪ Failure detection and prevention
▪ Scripted
▪ Limited process automation based on static rules
▪ Individual metrics based on structured data in lifecycle silo
▪ Limited insight and perspectives
▪ Reactive: deals with “Past”
▪ Failure prediction
▪ Automate the automation
▪ Intelligent process automation based on dynamic inference
▪ Analytics based on large scale structured and unstructured data across the lifecycle
▪ Predictive: from “What did happen?” to “What would happen?
TRADITIONAL TESTING
COGNITIVETESTING
Digital TestingIntelligent Analytics - Characteristics
Inspired by “Intelligent DevOps – Driving Continuous Delivery by Harnessing the Power of Analytics & Autonomics”, Shamim Ahmed, 11/16/2016, CA World 2016, Las Vegas
23
AUTOMATED & SELF-HEALING
DYNAMIC/ADAPTIVE RULES
BIG DATA ANALYTICS
IOT-Ops
DX-Ops SEC-Ops
INTELIGENT / ADAPTIVE / COGNITIVE
+ +
+
+
+
+
+
Digital TestingChallenges addressed by Cognitive Adaptive Delivery (CA-D)
Inspired by “Intelligent DevOps – Driving Continuous Delivery by Harnessing the Power of Analytics & Autonomics”, Shamim Ahmed, 11/16/2016, CA World 2016, Las Vegas
24 © 2017 CA. ALL RIGHTS RESERVED.
Digital Testing – Cognitive Adaptive Insight
25
Deep reinforcement learning
Digital TestingCognitive Adaptive Learning (CA-L)
Machine learning with neural dictionaries
https://www.luis.ai
“We are not at the centre of the universe (Copernicus), of the biological kingdom (Darwin) or of the realm of rationality (Freud). After Turing, we are no longer the centre of the infosphere, we share the infosphere with digital technologies”. The ethics of artificial intelligence, Luciano Floridi
26
Test Data Set
Test Algorithms Set
Test Meta Data
Source - IEEE Transactions on Pattern Analysis and Machine Intelligence, 21st September 2016, http://arxiv.org/abs/1609.06647Source - Google AI APIs - https://cloud.google.com/prediction/docs/apis
Digital TestingPattern analysis and machine intelligence
27
REAL TIME, BIG DATA ANALYTICS EMBEDDED THROUGHOUT OUR PORTFOLIO
AGILE MANAGEMENT AGILE OPERATIONS DEVELOPER PRODUCTS
CONTINUOUS DELIVERY MAINFRAME SECURITY
27
API analyticsRelease management dashboard Access incident responseReal-time app performance
Digital TestingContinuous Adaptive Delivery (CA-D)
Cognitive Adaptive Insight
Inspired by “Intelligent DevOps – Driving Continuous Delivery by Harnessing the Power of Analytics & Autonomics”, Shamim Ahmed, 11/16/2016, CA World 2016, Las Vegas
31
Digital Testing – Cognitive Adaptive Learning
32
Emergent embedded AIfor neural networks
Digital TechnologyCognitive Adaptive Intelligence (iCA)
Singularity Ultra-Intelligence (2030)
“Let an ultra-intelligent machine be defined as a machine that far surpass all the intellectual activates of any person however clever” 1960s
33
Mode 0 - Descriptive
Cognitive Adaptive Testing (CAT):
1. “How long will it take to test these new features?” --- How long historically (test cases / execution time)2. “What is the associated risk rating of the new platform?” --- Failure rate (cyclomatic complexity > 13)3. “How much technical debt do we have?” 4. “What is the defect density and prediction for this solution?”5. “What is the current conference rating of the solution?”6. “What is the value of this capability?”
How will Intelligent Analytics help?
Digital TestingExample use cases for Intelligent Analytics
34
What were the past defect trends?
What were the root cause of the past defects?
What defects are we likely to find?
What preventive measures should we take?
Have we overlooked anything for defect prevention?
Predictive analytics
Descriptive Analytics
Diagnostic Analytics
Prescriptive Analytics
Deep Learning
Diagnosis Analytics EngineExamples of Analytics use cases
Cognitive Adaptive Intelligence (CAi)
Inspired by “Intelligent DevOps – Driving Continuous Delivery by Harnessing the Power of Analytics & Autonomics”, Shamim Ahmed, 11/16/2016, CA World 2016, Las Vegas
35
How do you pin point the failures within the release
pipeline?
Model-Driven Design of Tests from Use Cases 1
2
3
Execute ‘Automation as Code’ from the Model
Pinpoint Failure and Diagnose Issue from Results
“Easy, soon as you check in the new features ready for test, the automation will run then the results are analysed automatically and if we have any failures then it identified them and raise them as bugs!
Cognitive Adaptive LearningTest in DevOps (TiD)
36
Mode 2 - Predictive
Cognitive Adaptive Testing (CAT):
1. “Can you create me all the tests I need to fully test this system?”2. “How do I successfully test this new feature?”3. “How much testing has been done already?”4. “Have you detected any anomalies within the system?”5. “How much testing will be enough?”
How will Intelligent Analytics help?
Digital TestingExample use cases for Intelligent Analytics
51
Where next?
54
Cognitive Adaptive Technology (CA-T)Next Generation of Digital Experiences (DX)
1. Digital Mobile, Chapter 2, Digital Experiences (DX), April 2017, Jonathon Wright
55
Q & A
58
Blogs #ExcuseFreeTesting – Legacy is your Legacy!
https://dzone.com/articles/legacy-is-our-legacy-1