AI-Assisted, AI-Augmented, &
Autonomous Testing
Choosing the Right Approach
Lee Vardaro Technical Solutions Architect @ Applitools
Agenda
✓ Defining the Landscape
✓ Core Components & Technologies
○ Coding Assistants
○ Test Case Generators
○ Test Code Generators
○ NLP Test Creation
○ Testing Agents
○ Visual AI for Testing
○ AI Analysis of Test Results
✓ Production Readiness
✓ Q&A
DEFINING THE
LANDSCAPE
AI-Augmented
Autonomous
AI-Assisted
✨ What are they?
✨ How do they differ from
traditional testing?
✨ What’s the role of human
oversights vs. full
automation?
DEFINING THE
LANDSCAPE ● AI engineering
● Prompt engineering
● Data security
● Cost of using AI API
In-House vs. 3rd Party AI
● Flaky tests
● Inconsistent results
Deterministic vs. Non-Deterministic
● The AI of today is not the AI of tomorrow
Rapid Change
CORE COMPONENTS & TECHNOLOGIES
● How they aid in
writing test-related
code
● Current maturity and
limitations in
production
Coding Assistants
● Automated creation of
test cases from logs,
user actions, and
design docs
● Real-world
effectiveness and
challenges
Test Case Generators
● Automation in
generating test scripts
● Balance between
speed and accuracy
● Test code not in actual
copy-paste format
Test Code Generators
CORE COMPONENTS & TECHNOLOGIES
● Converting plain language into
automated tests
● Defining individual steps vs.
intent driven testing
● Ability to record user actions
and convert to NLP
NLP Test Creation (LCNC)
● Autonomous/semi-autonomous
systems that mimic human testing
● Features like self-healing,
exploratory testing, and
AI-powered insights
● Production reality: benefits versus
areas needing improvement
Testing Agents
CORE COMPONENTS & TECHNOLOGIES
● Forecasting failures and
high-risk areas of applications
using data analytics
● Determining root causes of test
failure
● Gives an overall picture of the
health of both the application
and testing suite
AI Analysis of Application
Data & Test Results
● Types of tools for pixel-perfect
verification
● Current performance in
dynamic UI environments
● DOM snapshot vs. traditional
screenshots
Visual AI for Testing
Evaluating the Reality
PRODUCTION
READINESS
✨ Comparison to traditional
manual and automated testing
✨ Pros: Efficiency, scalability, and
cost-effectiveness
✨ Cons: Limitations in adaptability,
human intuition, and complex
scenarios
✨ Case insights: Where AI is
effectively deployed and where
it still falls short
Want to explore AI-powered testing
with Applitools?
Schedule a call with a
Testing Specialist
applitools.info/v2j
Questions??

AI-Assisted, AI-Augmented & Autonomous Testing

  • 1.
    AI-Assisted, AI-Augmented, & AutonomousTesting Choosing the Right Approach Lee Vardaro Technical Solutions Architect @ Applitools
  • 2.
    Agenda ✓ Defining theLandscape ✓ Core Components & Technologies ○ Coding Assistants ○ Test Case Generators ○ Test Code Generators ○ NLP Test Creation ○ Testing Agents ○ Visual AI for Testing ○ AI Analysis of Test Results ✓ Production Readiness ✓ Q&A
  • 3.
    DEFINING THE LANDSCAPE AI-Augmented Autonomous AI-Assisted ✨ Whatare they? ✨ How do they differ from traditional testing? ✨ What’s the role of human oversights vs. full automation?
  • 4.
    DEFINING THE LANDSCAPE ●AI engineering ● Prompt engineering ● Data security ● Cost of using AI API In-House vs. 3rd Party AI ● Flaky tests ● Inconsistent results Deterministic vs. Non-Deterministic ● The AI of today is not the AI of tomorrow Rapid Change
  • 5.
    CORE COMPONENTS &TECHNOLOGIES ● How they aid in writing test-related code ● Current maturity and limitations in production Coding Assistants ● Automated creation of test cases from logs, user actions, and design docs ● Real-world effectiveness and challenges Test Case Generators ● Automation in generating test scripts ● Balance between speed and accuracy ● Test code not in actual copy-paste format Test Code Generators
  • 6.
    CORE COMPONENTS &TECHNOLOGIES ● Converting plain language into automated tests ● Defining individual steps vs. intent driven testing ● Ability to record user actions and convert to NLP NLP Test Creation (LCNC) ● Autonomous/semi-autonomous systems that mimic human testing ● Features like self-healing, exploratory testing, and AI-powered insights ● Production reality: benefits versus areas needing improvement Testing Agents
  • 7.
    CORE COMPONENTS &TECHNOLOGIES ● Forecasting failures and high-risk areas of applications using data analytics ● Determining root causes of test failure ● Gives an overall picture of the health of both the application and testing suite AI Analysis of Application Data & Test Results ● Types of tools for pixel-perfect verification ● Current performance in dynamic UI environments ● DOM snapshot vs. traditional screenshots Visual AI for Testing
  • 8.
    Evaluating the Reality PRODUCTION READINESS ✨Comparison to traditional manual and automated testing ✨ Pros: Efficiency, scalability, and cost-effectiveness ✨ Cons: Limitations in adaptability, human intuition, and complex scenarios ✨ Case insights: Where AI is effectively deployed and where it still falls short
  • 9.
    Want to exploreAI-powered testing with Applitools? Schedule a call with a Testing Specialist applitools.info/v2j Questions??