Using Generative AI in Software Automation/Manual Testing (Course)

hi guys my name is Karthik and I am from a automation.com and welcome to my all new course in ex automation on generative AI in software testing especially automation testing and this

is the first of its kind course in generative AI where we are going to be talking about using generative AI as an application for software testing for

both manual as well as automator testing and this is going to be the first video where we are going to be talking about how generative AI is going to be

leveraging the power in the software testing field and this video is going to be an introduction to the course and we will understand how we can liver the power of gen in software automation

testing so why generative AI in software testing and what are the roles of artificial intelligence in software

testing and how will this AI or generative AI can help me get to help me understand how I can perform the software automation or manual testing in

a better way well generative AI or gen AI refers to a class of artificial intelligence model designed to generate

new original contents or data based on the input it receives it can create text image musics and even code which makes it more powerful across various field

and the one field which we are more interested is going to be the manual testing and automation testing field but how will this gen can help me in

software testing that is what we are more interested in and how will this gen AI generates manual test cases and automated test code if I'm going to be

looking at well in order for you to understand this entire operation of generative AI first of all we need to understand that generative AI by itself is not something out of the box going to

generate things for for you because you need to understand first how generative AI works by itself as you can see over here generative AI has got three most

important models as of today the first one is going to be the large language models or llms and then there is going to be a

Gans or Gans which is going to be a generative adversive networks and then vs or otherwise called as the variational auto encoder so these are

the three different models available under generative AI umbrella so these are the models or applications which are used within generative Ai and the large

language models are the one that we always hear for past couple of years right now the introduction of the chat GPT was the first thing which we know or

we encountered to use the large language models and we all wondered how everything is happening magically so you can see that the large language models actually has got three most important

purpose it has got text generation code generation and it also supports chat bots so you can see that as it has the name code generation it does all the thing that you're probably looking for

but before I get into large language models more deeper I also wanted to touch base with the Gans and the VAS and you can see that the generative adverse

networks has got uh option of image generation and also performs The Deep F verification and also it performs art creation or image creation and video

creation so it does all those operation for you over here and these are going to generate some new images that you're looking for most of the use cases that

you can see currently is the doll e is going to be mostly using the Gans of the generative Ai and then vs or otherwise

called as the variational auto encoders will also be generating new data but it also learns how to compress those data because you know that the guns are going

to be generating new images but what if the images and the video size is too heavy if you want to compress those images then you need to have the data compression and synthetic data

generation based on that and that's all going to be taken care by using this model the vas model so all these are included as a part of this generative Ai

and the one that we are going to be mostly focusing on is going to be the large language model models so what is this large language model as it names it is going to be a very very large

language model which is going to be helpful for generating one of the most important thing that we are going to be looking for is going to be the software CES because it does very good on that it

can understand and generate software code across various programming language it's not restricted to just one single programming language like C or Java or

python it also generates with Ruby r goang or any other programming language that you name it it just works fine it in fact work with more human languages

as well like Hindi Tamil Telugu marati Canada something like that so it all works quite fine with the large language model and this is happening because large language model is an application

of the generative AI as you know already and large language model encompasses quite a lot of different operation it it does the data extraction from various sources it gets all the input uses the

neural network to store all the information and it knows how to tokenize those Value store across it like a human brain and then it can retrieve the value that you're looking for based on the

number of token you already have and number of token it can accept so I'm talking about tokens and stuff in this particular lecture but I'm really not going to go into deep about the tokens and how the tokens are going to be

categorized in the new networks and stuff but yes large language model is the one which is responsible for you to generate the code that you are looking for well as that said there are many

popular large language models available and some of the most popular large language models today available in the market is the GPT 4 or 40 turbo we

already know that which is from open Ai and then we also have got llama from meta or otherwise called as Facebook so they have got a open-source version of

the large language models which means you can download the large language model within your own machine and then you can even run that offline if you wanted to and also there is something

called as Gemini which is from Google and this large language model is also very popular because it does quite a lot of operation pretty close to what GPT and llama does and there is another very

popular large language model which is the CLA 3.5 Sonet because this is from anthropic and it is very very good because it has got quite a lot of

different operation very unique compared to rest of these large language models as you can see over here because it also has got something called as artifact which generates things for you like a

separate file which you can easily view what that you're looking for and similar there is this mistal from mral AI it's also very popular large language model and every single large language model that you are seeing over here is all

available in some of the popular Cloud platforms like Azure Google and even in oracles AI if you wanted to so all these language models can be obtained from

there and you can just start using it within your organization or within your own personal hobby project and then you can liiz the power of large language model in testing as well as in

application development well as that said you might know all these information that I'm really talking about you may ask like Karthik what's really point I mean we all know about these things what is exactly that you are trying to talk about in this

particular course well we can see that there are many different applications of software testing within gen and that's what we are really going to talk about

there are popular applications of gen in software testing as well for instance if you're going to be very focused on the manual testing then gen AI can help you

generate manual test cases automatically it also does data mapping for you what does that really mean is that it tells you that whether this data can really

maps to that particular attribute of your applications data that you're looking for it does that and similarly you can also use this gen for performing automated software testing and code

Generations what does that really mean is that while you try writing any code in selenium or playright or any other tool AI will do the code suggestion and sometimes it also create a big gigantic

code for you that you can just use it to write the code on that and similarly it also does the AI code correction and code coverage which means it will help you correct the code while you try to

write and also it ensures that the code coverage happens based on the context of the entire application that you are trying to test because you can also these days pass the entire source code

of your application as a context to the generative AI tools for instance GitHub co-pilot or tab 9 and then you can use

them to be using as a context for writing the automated software testing tool and you can also use this jni to generate test data automated bug

tracking which can also be helpful if you're going to have a duplicate bug already been rised on that similar content so those things can even happen with these generative AI tool and one of

the most popular example of that is the busura from Mulia software and that already has this automated Buck tracking in it so these are very very popular and

there are few more use cases that you can just think of with this generative AI in automation testing and manual testing is that if you have got a big swager documentation with all the API

definitions and you wanted to write an API testing with Postman or restar then gener a AI can do it for you automatically what does it really mean is that just copy paste that in any of

these large language models it will give you the suggestion for what exactly that you are trying to test and it will give you the information that you're looking for these are going to be very very

handy because you just have to have a bit of a walkth through on the cord which is been generated or maybe just do a bit of an inspection whether the code that you're looking for is exactly what

is there the all the babysitting work that you have been doing doing all these days will be gone because it's all going to be taken care by this gen and also there is another use case that if you

got a fragile UI application which happens most of the time while we do a UI application testing then all you have to do is you just have to pass the code with the page object model or whatever

to the Gen and ensure that the locators that you have got is matching with the page that you have got that you can just pass the source of the page ensure that the locators are always matching if it

is if it's returning you the True Value then you can just ensure that the test just works fine and if there is a failure happens it is not because of the application uh UI change rather it is a

logic change so you can do that as well it is so good with this particular option but you need an API options like

you need to have an API of the Genna like chat GPT or maybe CLA something like that and then you can perform those operation but at least yeah you can do

those things so this is another use case and third use case is very very popular and I've seen that many people are really using it is generating test data for complex object in more realistic

fashion because we have seen many time while you wanted to generate or probably pass in the data for the application you may have seen that you need to ensure

that the test data is not duplicated and also more realistic and also you don't really babysitting working on the same test I entry every single time this gen

can help you do that for you and it reduces quite a lot of time and you can see that all these use cases that we have seen so far is just trying us to

help to make our life more easier to perform all the testing as well as the babysitting work away from us and that

is the more powerful thing about this generative Ai and that's exactly what we are going to be covering in this entire course stacking each and every use cases

in more Enterprise grade realistic fashion and we'll see how we can get through this course in a more meaningful way to achieve all the operation using

generative AI both offline and online so let's learn this entire course and understand how things work in the world of generative AI I'm so excited with this course thank you so much for

watching this video catch you starting our first section of this course