How AI is transforming Software Testing

schedule Jun 29th 11:30 AM - 12:15 PM place Lalit 1 people 94 Interested

About 5 years ago everyone was talking about “Mobile First” and giving the user a mobile experience using mobile web, native and hybrid applications. Now, the new buzzword is AI. Not a day goes by without people or articles mentioning some sort of AI development happening in the industry. It is everywhere, in self-driving cars, speech recognition (people sure do love their Alexa :-)), computer vision, healthcare, fintech and now in software testing. Artificial Intelligence and machine learning are advancing at a rapid pace. Companies like Apple, Tesla, Google, Amazon, Facebook and others have started investing more into AI to solve different technological problems in the areas of healthcare, autonomous cars, search engines, predictive modeling and much more. Applying AI is real. It’s coming fast. It’s going to affect every business, no matter how big or small. This being the case how are we as Testers going to adapt to this change and embrace AI? Also, in the era of where we want to automate everything, how is AI going to influence the way we do test design and automation? Come join this session, where I cover the basics of AI, discuss the key ways test automation can benefit from AI and the challenges involved in implementing AI based solutions. Attending this session will help anyone to get started with AI based testing.

 
 

Outline/Structure of the Talk

Talk Outline

Learn the basics of AI

  • Quick intro on what it is
  • Difference between AI, Machine Learning, Deep Learning
  • What is Deep Learning
  • Supervised Learning with real examples
    • We give the right data for the algorithm to learn
  • Unsupervised Learning with real examples
    • We give a bunch of data and see what relationships you can find
  • Reinforced Learning with real examples
    • Concept of Reward Function. Reward Good/Bad Behavior
    • E.g. Training a Dog
  • Ask people to use an application with AI for a quick interactive exercise and show example of AI in real application
    • Quick Draw from Google

Challenges with Software Testing Automation

  • Discuss challenge of Skill set, authoring initialization, maintenance and scale

How AI can help to solve these challenges

  • Discuss Dynamic Locators and AI with Demo

Future of AI in software testing

  • Discuss different ways AI is being used and will be used in the future

Learning Outcome

Attendees will get to know the basics of AI, how to test AI based applications and how it is influencing software testing

Target Audience

Anyone interested to know the basics of AI and how to test AI based systems

Prerequisites for Attendees

Nothing needed

schedule Submitted 1 year ago

Public Feedback

comment Suggest improvements to the Speaker
  • Anand Bagmar
    By Anand Bagmar  ~  1 year ago
    reply Reply

    Hi Raj,

    This is an interesting topic. Can you share some draft slides for the program team to understand the content in a little more detail?

    • Raj Subramanian
      By Raj Subramanian  ~  1 year ago
      reply Reply

      Talk Outline

       Learn the basics of AI

      • Quick intro on what it is
      • Supervised Learning with real examples
        • We give the right data for the algorithm to learn
      • Unsupervised Learning with real examples
        • We give a bunch of data and see what relationships you can find
      • Reinforced Learning with real examples
        • Concept of Reward Function. Reward Good/Bad Behavior
        • E.g. Training a Dog
      • Some of the real examples of AI I discuss with videos would be
        • Alpha Go, Deep Blue
        • Recommendation engines on Amazon, Netflix where customer is recommended products and movies
        • Amazon uses AI for logistics 
        • Car Optimization
        • Identify which e-mail is spam from group of e-mails
        • extracting text from audio
        • Loan applications apps where AI determines how probable a user would be able to pay the loan and grant loan accordingly
        • Autonomous cars
      • Give an example of how AI could work (Attached an image as an example of AI being used to predict price of the house based on the size)
        • You keep giving inputs such as bedrooms, zipcode, wealth etc to see how it affects prices of houses
      • Ask people to use an application with AI for a quick interactive exercise and show example of AI in real application
        • Quick Draw from Google

       

      How to test AI based systems

      • Giving Intelligent and Valuable Training Sets to train the AI. Will have examples
      • Monitor how the AI system is learning and track its progress
      • Keep beta testing with other user interaction
      • Keep collecting large amount of data
      • Have sufficient computing power to process all this data

       

       How software testing can benefit from AI (This will have live demos)

      • Will focus on Automated Testing
      • Discuss state of automation before AI
      • How AI can be used in automation tools 
        • Discuss how Testim.io uses AI with live demos
          • Will Discuss how  the Dynamic locators work in real time (They get recorded in the tool for the user to see). How it makes the test stable
            • Will show a live example of how changing the attribute of an element on a web page won’t break our tests because of the AI.
            • I will change the name of  a button say “Submit” to “Log in”. Change its id, location on the screen, position on the screen
            • Then will run the same exact test without any re-recording and our AI will detect this change and still keeps continuing to run the tests as it locates other ways to identify the “Submit” button
        • Discuss how other vendors are using AI for software testing with live demos

       

      Identify and discuss challenges involved in implement an AI based solution

      • Discuss Data and Computing the biggest challenges of AI with examples
      • Future of AI on how it will be used in software testing
        • Finding Production and Monitoring issues
        • Logs
        • Give a lot of screenshots and find patterns
        • Get automation to a state where it can be cut into different pieces and run in parallel. Saving time
        • Identifying code problems based on analyzing server requests etc
        • The user flows could be created by production users
        • Finding Churn percentage of uses
        • Learn to use adequate Wait's between pages
      • Why do we need to harness the power of AI (Discuss various use of AI in the present and future)
        • Speech recognition
          • Accents
          • Low bandwidth areas
          • Cafe
          • Voice Assistant
      • Computer Vision
        • Face Recognition
        • Healthcare
        • Medical Imaging
        • FinTech
      • Discuss Weka -> Open source project for Machine Learning
        • Will have a demo of how people are using it
      • Naresh Jain
        By Naresh Jain  ~  1 year ago
        reply Reply

        Thanks, Raj. Can you please update your proposal instead of putting the outline in the comment.

        Also, you can share any slides or videos from your past presentation. Does not have to be from the current one. This helps the program committee understand your presentation style and make an informed decision.

        • Raj Subramanian
          By Raj Subramanian  ~  1 year ago
          reply Reply

          Just did. It is my bad that I did not update it properly in the first place. I apologize for the confusion.

          • Anand Bagmar
            By Anand Bagmar  ~  1 year ago
            reply Reply

            Hi Raj,

            Thanks for adding the details.

            Only concern I have is that you are attempting to cover a lot of information, which is going to be new I guess for a good portion of the attendees. Will you be able to go through all the items listed, in decent time and with appropriate speed? Also, for the demo, do you have any specific requirements?

            • Raj Subramanian
              By Raj Subramanian  ~  1 year ago
              reply Reply

              Hi Anand,

              I gave you more detailed perspective on what content I was exactly trying to cover so that it will be easy for your committee and people reading my proposal to know what I will be covering during the session. I will not be going over each and every example on the intro section (just listed everything I can think of for your reference) and would be spending more time on the demo and how to test AI based systems.

              This way everything I cover should fit within the allotted time. The way I will be spending my allotted time of 45 mins would be

              Learning the basics of AI (5 mins with a video)

              How to test AI based systems (5 mins)

              How software testing can benefit from AI (25 mins with demos)

              Challenges and future of AI (5 mins with a video and a demo)

              5 mins buffer time based on audience interaction

              I will be making the presentation as simple as possible to understand geared towards first timers in this space. Believe me I myself in the learning stage and it takes a lot of reading to get an idea about it :-)

              As for requirements for the demo I just need a projector and screen. I would also require audio connection to show couple of videos related to AI.