Machine Learning Heralds the End of Selenium Presented by: Jason - - PDF document

machine learning heralds the end of selenium
SMART_READER_LITE
LIVE PREVIEW

Machine Learning Heralds the End of Selenium Presented by: Jason - - PDF document

W10 Test Analytics, AI/ ML Wednesday, May 2nd, 2018 1:45 PM Machine Learning Heralds the End of Selenium Presented by: Jason Arbon Appdiff, Inc. Brought to you by: 350 Corporate Way, Suite 400, Orange Park, FL 32073 888 --- 268 --- 8770


slide-1
SLIDE 1

W10

Test Analytics, AI/ ML Wednesday, May 2nd, 2018 1:45 PM

Machine Learning Heralds the End of Selenium

Presented by:

Jason Arbon

Appdiff, Inc.

Brought to you by:

350 Corporate Way, Suite 400, Orange Park, FL 32073 888---268---8770 ·· 904---278---0524 - info@techwell.com - http://www.stareast.techwell.com/

slide-2
SLIDE 2

Jason Arbon

Appdiff, Inc.

Jason Arbon is the CEO of Appdiff, which is redefining how enterprises develop, test, and ship mobile apps with zero code and zero setup required. He was formerly the director of engineering and product at Applause.com/uTest.com, where he led product strategy to deliver crowdsourced testing via more than 250,000 community members and created the app store data analytics service. Jason previously held engineering leadership roles at Google and Microsoft and coauthored How Google Tests Software and App Quality: Secrets for Agile App Teams.

slide-3
SLIDE 3

4/25/18 1

Machine Learning Heralds the End of Selenium

Jason Arbon

#STAREAST #AI @jarbon @testdotai

slide-4
SLIDE 4

4/25/18 2

#STAREAST #AI @jarbon @testdotai

slide-5
SLIDE 5

4/25/18 3

so 2004

#STAREAST #AI @jarbon @testdotai

4 Big Problems with Selenium / Appium

#STAREAST #AI @jarbon @testdotai

slide-6
SLIDE 6

4/25/18 4

You Have To Code

#STAREAST #AI @jarbon @testdotai

Element Selection

#STAREAST #AI @jarbon @testdotai

slide-7
SLIDE 7

4/25/18 5

Hard-Coded Step Sequencing

#STAREAST #AI @jarbon @testdotai

Little Reuse

#STAREAST #AI @jarbon @testdotai

slide-8
SLIDE 8

4/25/18 6

Automation breaks just when you need it.

#STAREAST #AI @jarbon @testdotai

Why AI?

#STAREAST #AI @jarbon @testdotai

slide-9
SLIDE 9

4/25/18 7

People can deal with change and the unexpected.

#STAREAST #AI @jarbon @testdotai

Test code can’t.

#STAREAST #AI @jarbon @testdotai

slide-10
SLIDE 10

4/25/18 8

How does AI fix all this?

#STAREAST #AI @jarbon @testdotai

AI works like People

#STAREAST #AI @jarbon @testdotai

slide-11
SLIDE 11

4/25/18 9

You Have To Code

#STAREAST #AI @jarbon @testdotai

Approach #2

Use Machine Learning directly instead of Selectors.

Approach #1

Use Machine Learning to smartly update/repair Selectors when they break.

#STAREAST #AI @jarbon @testdotai

slide-12
SLIDE 12

4/25/18 10

Approach #1 VS

Element Identification

Test.ai Auto Labeled Quick Visual Custom Labeling Selenium IDE Pain Finding Identifiers Have to write code (brittle)

Old / Past

slide-13
SLIDE 13

4/25/18 11

Element Selection

#STAREAST #AI @jarbon @testdotai

slide-14
SLIDE 14

4/25/18 12

slide-15
SLIDE 15

4/25/18 13

Hard-Coded Step Sequencing

#STAREAST #AI @jarbon @testdotai

VS

Step Sequencing

Selenium IDE Have to code exact steps. Brittle when app changes. Test.ai Familiar Cucumber/Gherkin Only list labels of interesting steps and verifications. Also, GUI Editor

Old / Past

#STAREAST #AI @jarbon @testdotai

slide-16
SLIDE 16

4/25/18 14

#STAREAST #AI @jarbon @testdotai

VS

Test Execution Old / Past

#STAREAST #AI @jarbon @testdotai

slide-17
SLIDE 17

4/25/18 15

Little Reuse

#STAREAST #AI @jarbon @testdotai

AIT - Open - Manual Too

#STAREAST #AI @jarbon @testdotai

slide-18
SLIDE 18

4/25/18 16

Write Once, Run Cross Platform

#STAREAST #AI @jarbon @testdotai

Game-Changer: Write Once, Run on All Similar Apps

#STAREAST #AI @jarbon @testdotai

slide-19
SLIDE 19

4/25/18 17

Many Companies Working on AI for Testing

#STAREAST #AI @jarbon @testdotai

Harnessing Collective Intelligence For Testing

slide-20
SLIDE 20

4/25/18 18

#STAREAST #AI @jarbon @testdotai

Leverage AI Approach #3

Record and look for Anomalies

Approach #4

Drag and Drop test

  • creation. AI executes tests

like a human.

Approach #1

Record and auto repair selectors

Approach #2

Visual Differencing Cross Platform

#STAREAST #AI @jarbon @testdotai

slide-21
SLIDE 21

4/25/18 19

Approach #1 Approach #2

slide-22
SLIDE 22

4/25/18 20

Approach #3 Approach #4

slide-23
SLIDE 23

4/25/18 21

Approach #4 VS

Test Reporting

Selenium IDE Many tests flaky/break as app changes Simple pass/fail Test.ai Q-Learning test step generation Element Classifiers at runtime See test steps Presentable

Old / Past

#STAREAST #AI @jarbon @testdotai

slide-24
SLIDE 24

4/25/18 22

#STAREAST #AI @jarbon @testdotai #STAREAST #AI @jarbon @testdotai

slide-25
SLIDE 25

4/25/18 23

jason@test.ai @jarbon

#STAREAST #AI @jarbon @testdotai