Dinakar
Uh let's quickly get started uh and so hello everyone and welcome to today's P cloudes webinar on enhancing test reliability how can AI help prevent test failures my name is dinaka and I'll be.
Dinakar
The host for today before we get started I would like to go over a few housekeeping rules your line is currently muted.
Dinakar
However you can submit your questions during the webinar using the Q&A option uh given in the bottom of the screen we will answer your questions as part of the Q&A section at the end.
Dinakar
Of the webinar and also please do note that this webinar is being recorded and uh we will send out the link of the recording to share with your colleagues or to watch again later.
Dinakar
With that said let me introduce the speaker for today Shueb Ahmed Shueb is a pre-sales manager at Pcloudy he has over 11 plus years of experience in the IT industry and prior to.
Dinakar
His role at Pcloudy he has worked with companies like microfocus as a pre-sales consultant uh he had done this in a application delivery management portfolio catering to different Technologies such as devops automation.
Dinakar
Performance cloud and Mobility so we have an expert here talking to us about how to make our Tests reliable.
Dinakar
So over to you Shueb, without further ado, let me hand it over to you. All right.
Shueb Ahmed
Thank you so much Dinakar for the introduction. Hello everyone, very warm welcome to all and uh uh.
Shueb Ahmed
Thank you so much for your time so today's topic is more around the test uh reliability so we will.
Shueb Ahmed
Look into various different factors of test reliability and uh how how we can enhance this with the help of AI all.
Shueb Ahmed
Right and prevent the test failures so without further delay let me uh get the agenda here is the uh agenda.
Shueb Ahmed
So when we talk about the agenda so here is a quick agenda so let's first we'll focus on what is test reliability uh in in in terms of you know the generic Bas and.
Shueb Ahmed
Then also we will try to understand in the software field in today's automation sector and uh also we will understand how the test becomes flacky how the test becomes flacky in nature and we.
Shueb Ahmed
Will also try to you know understand or know try to overcome these issues that arises because of lackness in the test uh we will.
Shueb Ahmed
Look into at various different options and then the very important the self feeling AI feature how this self feeling AI can help overcome the flakiness and.
Shueb Ahmed
Also overcome the test and reliabilities and then we will also look into some some of the advanced test analytics.
Shueb Ahmed
Right so there are various different analytics that we uh as a platform provide so how they can help in overcome the the different challenges that arises and.
Shueb Ahmed
Then followed by your live demonstrations at the end we will have a Q&A session so in case if you have any questions feel free to put your questions uh in the chat I'll be.
Shueb Ahmed
Happy to take it at the end of this uh demonstration all right so with that let's get into action and about the reliability.
Shueb Ahmed
So when we talk about test reliability what exactly it is so if you talk in general generic terms test reliability refers to an extent that it measures a test measures uh without any errors.
Shueb Ahmed
Right so without any errors is basically term it as a test reliability so we can we can consider this test reliability in various different sections.
Shueb Ahmed
Right but we can take it in the software field so if we talk about in today's U software testing.
Shueb Ahmed
Right test reliability is to ensure and test the software that it works without any issues right and same case we can apply it to all our automation test scenarios as.
Shueb Ahmed
Well right so in general if you talk about in today's uh quick evolving technology sector the usage of software application has been increasing rapidly daytoday in our lives.
Shueb Ahmed
So be it at telecommunication sector or even fintech or even from the healthcare side uh e-commerce or even the government sector everywhere we are you know we are working with the softwares everywhere software.
Shueb Ahmed
Is involved so it is very important to test our applications so that you know they can be reliable right.
Shueb Ahmed
So there are various different types of test that are conducted right so starting from manual testing to automation testing uh performance testing.
Shueb Ahmed
So there are various different types so the very important p is to understand how the test can be reliable.
Shueb Ahmed
So there are certain factors where the test results are not dependent all right uh so when we talk about the OCC.
Shueb Ahmed
So you test it with on any time anywhere the test results will Still Remains the Same so this is when when we talk about when the tests are reliable and.
Shueb Ahmed
Also when we talk about any feature specific you test any uh any test suit any part of a test you pull out a test case you pull out a test case and you test.
Shueb Ahmed
The test will Still Remains the Same this is also one of the criterias where we say the test are reliable and.
Shueb Ahmed
Also it's not dependent on any tester anyone test the software or anyone test the applications it should be Flawless that's where we say the test is reliability test reliability is very important in nature.
Shueb Ahmed
Now what affects this test reliability there are various factors I get unreliable test will lead to inconsistent results it can be uh the failures the number of pass rate versus the failure rates uh.
Shueb Ahmed
Could be again there are various different factors um the why are the tests failing because of inconsistent in code it can be.
Shueb Ahmed
Because of the environment or because of the lack of test and this will further lead to flacky test so let's try and understand what flacky test are in general.
Shueb Ahmed
Right so when we talk about uh flacky test so flacky test in general it refers to the testing which generates uh you know inconsistent results uh you know could be from failing or passing.
Shueb Ahmed
And unpredictability so flak test is will certainly lead to lot of you know issues within the applications and the automation test scenarios that we are testing and that that again leads to unreliable test.
Shueb Ahmed
So here is one quick survey uh by uh two of Jane software in the world so as part the Google it says that the flacky test accounted for about 16% you know that is.
Shueb Ahmed
You know of all the test failures in their system so that's something which is hugee about 1.5 times longer time.
Shueb Ahmed
So when we talk about uh you know how do we identify uh you know finding these tests it can take a longer time than to fix the non- flaged test ones.
Shueb Ahmed
So that's you the 16% is not a small rate right and also when we there's another survey from Microsoft which talks more on the revenue site all.
Shueb Ahmed
Right which states that you know about 1.14 billion per year in terms of the developer time is being spent.
Shueb Ahmed
Right so that's an estimate uh that you know the flag test has been leading to so that will add to the revenue hit and again um know there's lack of abandoning from the user.
Shueb Ahmed
Standpoint how do these test become flaky in nature right so there are again various uh different factors so we have listed uh few of the uh most um you know consistent uh factors that.
Shueb Ahmed
Become that can lead to a flest so one is the external elements like the apis so so interaction with an external systems like the uh apis or even the databases that introduces flakiness and.
Shueb Ahmed
Uh due to the factors could be from the network latency standpoint even the um wearing response time so this is one of the important factor which is leading to a flaky test and.
Shueb Ahmed
Also when we talk about the environment side unstable environment so so when the test environment is isolated not controlled or even consistent.
Shueb Ahmed
So this will lead to variations in the performance and as well as the availability uh on on the configuration side.
Shueb Ahmed
So it's very important to have a stable environment and the next the next pie is are the uh test data.
Shueb Ahmed
So if the if the test data is in dynamic in nature so this can also lead to uh know flaky test.
Shueb Ahmed
So it's very important to handle the dynamic data so now that we are working with complex applications in the in today's time all.
Shueb Ahmed
Right so the data is very much Dynamic so we cannot have every time a static data so that in order to make the know the FL it test to be reduced.
Shueb Ahmed
So there are there there is a complex scenarios there are complex applications built on multiple uh Technologies and so we have a dynamic set of data.
Shueb Ahmed
So it is very important to handle these Dynamic set of data properly and with more Dynamic set of data there are chances of having more flakiness within the test and.
Shueb Ahmed
Then of course logical flaws so when the test contains the bugs uh the typos or even the errors uh errors in the code.
Shueb Ahmed
So this can lead to a logical fla and then and the last one is that one which we have highlighted is insufficient checks if if the test has not gone through uh sufficient checks.
Shueb Ahmed
Then this can also become uh lead to having more flecky test in nature and there are of course few other dependencies um I mean it could be at the application side and even uh.
Shueb Ahmed
You know the elements uh the elements within the applications like the dates the times the uid the user inputs there are various different factors.
Shueb Ahmed
So these are the most common um attributes are the factors that are leading to make the test flaky now how do we overcome this mess.
Shueb Ahmed
Right that's untangling the mess so it's very important to understand the failure what led to the failure of the test identify the uh the failures with the proper reports.
Shueb Ahmed
Right and also identify the root cause so is my platform that I am using or is my automation tool that I using is helping me to investigate the failure is something which is very.
Shueb Ahmed
Important so investigating the failure is really important to overcome this issue and whatever test that you are performing you have to make sure that if it is valuable or not if it is valuable.
Shueb Ahmed
Go ahead and then fix it if not just go ahead just delete and move away this otherwise there's a lot of time being spent again in identifying uh what cost and there's unnecessary effort.
Shueb Ahmed
That's been spent for a test which is not valuable right it's always better to move on uh if you have able to identify the fix just move on and.
Shueb Ahmed
Then associate these failures and tag them uh appropriate if you have identified the failures tag it properly so that we can.
Shueb Ahmed
So that whosoever is working from the developer standpoint they can understand uh what are all the failures and U and last but not least is to check if uh the problem is within the.
Shueb Ahmed
Test or even the environment so environment can also play a huge role as we saw in the previous slide.
Shueb Ahmed
So it's very important to determine what led to a problem is it because of the test itself or an environment most of the time what happens is we ignore more on the environmental factors.
Shueb Ahmed
Uh environment side um so environment side is something which is very important we need a stable environment where the uh the test execution takes place not every time it can be test of course.
Shueb Ahmed
Test there can be many flaky test but environment can also contribute right so here is the um self feeling how self feeling can help overcome these challenges all.
Shueb Ahmed
Right so it is self feeling is basically um you know a feature wherein helps you heal your automation scripts um you know without failures.
Shueb Ahmed
So that is something which can help you from the maintenance standpoint of your test so if you look at it on an um at a higher level.
Shueb Ahmed
So it potentially identifies the failures so which is very important so while you're execution is going on it's very important to identify where exactly the issues are.
Shueb Ahmed
So that's number one and then it fixes the flacky scripts right if there are test that are written and they are flacky in nature and the reasons we have we have looked in our.
Shueb Ahmed
Previous slides so it helps fixes them so that's something which is very important so we will look into detail in the live demonstration I I have a live demonstration which can be discussed and.
Shueb Ahmed
Then it automatically updates all the elements and the object locators you don't have to spend time in updating the elements or the locators.
Shueb Ahmed
So there are possibilities that in your application uh when you have created a test and when you have a new build all together there are chances that the object elements could be changed.
Shueb Ahmed
Right and spending effort in identifying these elements and locators takes a lot of time that's where um you know the test maintenance effort goes very high.
Shueb Ahmed
Right so the this self feeling AI feature can help you overcome the uh the challenges that are possessed from the flaky and.
Shueb Ahmed
Also it will update all the elements and the object locators automatically so that your test runs without any failures and.
Shueb Ahmed
Yeah as I mentioned I mean uh this will further effort the uh reduces the maintenance effort you don't have to spend much time if the platforms are built with the self feeling um you.
Shueb Ahmed
Know capability it can automatically heal your scripts the failures will be taken care if there are any fixes within the scripts will be taken care and.
Shueb Ahmed
Also from the El standpoint and and net overall if you look at it on a broader picture it will reduce a lot of effort which in terms helps you in overcoming the uh cost.
Shueb Ahmed
And so on and we can also talk about the analytics so analytics is very important piece all right so realtime analytics is something which is very important.
Shueb Ahmed
So we will look into what are all the parameters that we provide in terms of um providing the realtime insights uh.
Shueb Ahmed
So insights into the real time metrics helps you you know in the faster feedbacks and also helps fixes you the issues you know before arising it.
Shueb Ahmed
Right so for your you know could be from uh know creating a truly continuous testing me ly so there are various different parameters again within the the test reports that has been provided.
Shueb Ahmed
So that can help analyze and quickly fix them and also it's very important to note the failure and the pass rate.
Shueb Ahmed
So so generating metrics such as the pass versus fail the frequency of the retires and also the time resolution to take to fix them.
Shueb Ahmed
So that's something which is very important and uh we will look into it and further last to it what is the impact of uh these analysis how do we um go ahead with this.
Shueb Ahmed
Analysis and help resolve the issues so that's something which we will be U uh looking in today's demon ation.
Shueb Ahmed
So not not much I'm not spend much time on the uh presentations but rather I'll try to focus more around the uh demonstration.
Shueb Ahmed
So let's uh get into action I'll take you through a live demo on our platform so just to give a quick overview on our platform.
Shueb Ahmed
So Pcloudy is uh basically the unified app testing Cloud platform where in we provide uh an automation solution inbu solution which is.
Shueb Ahmed
Basically codess from the codess automation standpoint where you can create your scripts and also execute them and we also have our own test infrastructure which comes up with its own uh physical devices that.
Shueb Ahmed
Are hosted across different part of the world uh not just mobile devices but also on the web browser instance if you're looking to test your web applications.
Shueb Ahmed
So you can seamlessly test and also uh our solution comes up within built test management solution so that's something uh where we term it as a unified platform.
Shueb Ahmed
So that becomes an end to end uh single test solution uh for your functional testing need replacing the fragmented tool chains.
Shueb Ahmed
So today we'll focus more around the uh the self feeling feature how it can help so I have two different builds um different builds just to give you an example.
Shueb Ahmed
So one build is basically uh you know know where we have created an automation scripts for a specific build and.
Shueb Ahmed
Now when we bring some changes in the new build how my AI uh self feeling feature can help overcome the issues.
Shueb Ahmed
So let me just quickly take you maybe I'll just quickly connect to a physical device on the cloud which is available these are real physical devices that are hosted on our Cloud I'm just.
Shueb Ahmed
Using it just to show you I'll just try to install one application and U try to get the objects and elements of that particular app as I mentioned.
Shueb Ahmed
So we have different builds install both the builds so that uh we are all on the same page and understand what was the change.
Shueb Ahmed
So these are the two builds one is the blood bank one application.
Shueb Ahmed
APK and the other is APK so I'll first try to install the first application and I'll get into the object spy mode.
Shueb Ahmed
So yeah the application has been installed so we have created an automation script for this particular scenario where we are launching the app signing in with the email ID and the password and.
Shueb Ahmed
Then signing in all right so that's the scenario very basic steps that we have performed and uh yeah we are executing this through an APM script.
Shueb Ahmed
Okay so if you here look at it this is my script yeah logging in with an username and this is my resource ID all.
Shueb Ahmed
Right and then followed by uh password and then we're logging in okay that's basic step what we have and we have configured it uh.
Shueb Ahmed
So that you know we can execute it on the Lo platform so let me go back and try to expect I'll just get into the object spy mode.
Shueb Ahmed
So here is an object spy mode so object spy mode is basically helps you identify the objects and elements uh and it helps you provide all the object hierarchy uh for the script creation.
Shueb Ahmed
Purpose and also gives you various different attributes and values so that's something which is an inbuilt feature that comes uh with from the platform perspective.
Shueb Ahmed
So as a first step I'll quickly go ahead and then click on start okay once I click on start the device in the the app will uh be ready for me to inspect.
Shueb Ahmed
Yeah now it's ready now I can just over my mouse over the fields various different fields so for example if I have to click on email I'll just click on email here and I'll.
Shueb Ahmed
Get the uh the hierarchy of this particular option right so if you look at it here so this is the resource ID this something which we can use it for my script creation purpose.
Shueb Ahmed
So there are various different options so which you can use it depending on um what is convenient to you and if you.
Shueb Ahmed
Look at so this is my resource ID com.
Shueb Ahmed
Android blood ID input username and same I'll just over the mouse over here so that you can have a.
Shueb Ahmed
Look at it so this is my resource ID which is something which I can copy okay something which I can use it in my Des scripts.
Shueb Ahmed
Now if I just go back to my script and show you here it is so this is the resource ID of my.
Shueb Ahmed
First build okay similarly I have a extracted for password and also for clicking on login button all right so whatever that I have captured is have translated into I have taken it to my.
Shueb Ahmed
Automation script now let me go back and also maybe if I have to show it for password I'll just click on password Here is the one it says input underscore password here I'll just.
Shueb Ahmed
Over the over here it says in input password yeah so that's the second field where we have got the resource ID and the last field is of course the login button.
Shueb Ahmed
Okay so this is my flow if I try to execute this script will work without any issues right that that's the typical Behavior.
Shueb Ahmed
Now what I will do is I'll go ahead and stop this and install another bu okay which I already have with some changes.
Shueb Ahmed
So we have brought some changes intentionally to this application so that to see if the platform can handle from the self filling standpoint.
Shueb Ahmed
So here is the second application blood bang.
Shueb Ahmed
APK I'll just go ahead and install it I I'm still in the object spy mode itself so that I can show you what Els have been changed.
Shueb Ahmed
Okay I'll just go ahead and click on email right so now if you look at it there is my resource idid has changed from input uncore username to just user if you.
Shueb Ahmed
Look at it this is the resource IDE I can just show it over here as well it says idea SL input user.
Shueb Ahmed
So just go back and just show you so my test step Remains the Same right my ID is changed.
Shueb Ahmed
Now so what I am doing is I'm not going to do this change so typically without self feeling feature what you tend to do is if your test scripts are failing.
Shueb Ahmed
Because of object properties getting changed you come here and change the resource ID that's the typical um uh you know scenario or the way of doing changes within your script.
Shueb Ahmed
Yeah same Cas is with password as well just click on password yeah is idpwd so it is instead of uh input password it has changed to id.
Shueb Ahmed
PW so here are the changes and so even for the login button so we have intensely changed all the different uh different parameters.
Shueb Ahmed
So that we can see whether this has changed so this is for id/ button.
Shueb Ahmed
Log and here it is id/ button.
Shueb Ahmed
Login okay I'll not this changes so now I have shown you the bus I'll just come out of the object spy mode.
Shueb Ahmed
So that I can go ahead and then execute it so let me release this device all right and now just go back to my existing script I'll just quickly um run through what all.
Shueb Ahmed
The parameters that we have used there is a runner.
Shueb Ahmed
Java file where we are uh initiating the uh driver okay and also driver.
Shueb Ahmed
Java this is where where we have Incorporated all my key capabilities okay for my application so first what I'll try to do is I have two different application blood bank one and blood bank.
Shueb Ahmed
2 so this script is designed to work with bank one okay because that's the resource ID what we have taken in the uh uh you know for our test cases.
Shueb Ahmed
Right so this is this is for the bled Bank application and uh yeah I have parameterized it the username and password.
Shueb Ahmed
So here's the um capabilities which we have used to you know integrate with the p platform just simple capabilities with stocks which.
Shueb Ahmed
Basically are the username P ke the application details followed by on which device you are going to execute okay.
Shueb Ahmed
So that's something which you can provide it here and the so this is since is Android based application so we are going with an Android and uh.
Shueb Ahmed
Yeah so these are the capabilities and I have configured my environmental as well just to provide the username and the password all.
Shueb Ahmed
Right so now we are ready now we can go ahead and execute so before that just let me quickly check ifel to is available just go here see still I can see both there.
Shueb Ahmed
Are two devices aailable I can go ahead and execute it remember this is blood bank one okay so now the just go to run.
Shueb Ahmed
Java and then quickly run as go the test okay the execution has been began and just go to the P platform the my active session I should be able to see the device booked.
Shueb Ahmed
Yeah the device is booked we'll wait for it to to get started so once this gets ongoing in the ongoing progress it'll take you through the uh device it's ongoing I'll just quickly connect.
Shueb Ahmed
Head by side I can monitor the logs as well yeah the application is installed it will pass as expected.
Shueb Ahmed
Because the resource ID whatever is taken is for the blood bank one application okay the execution got uh did without any issues all.
Shueb Ahmed
Right so now let's take a look at the results yeah username entered password entered login click demo Pass Plus these test are based on resource ID which we have taken for the application build.
Shueb Ahmed
Number one okay so it it remains perfect same now for the new build which we have that is nothing but blood bank to which I have just which I just installed and showed you.
Shueb Ahmed
That resource ID has been changed for the username password and also for the login button so I'm not doing any change here.
Shueb Ahmed
So what I'll try to do is I'll just go here and then change the application name from one to.
Shueb Ahmed
Okay I've just changed it save it and I'll just come back I can run it what I'll do is I'll just release this device maybe if you if I want I can just run.
Shueb Ahmed
It on any other device as well let's see if uh Galaxy F14 is available I'm just taken it note over here just go back.
Shueb Ahmed
Yeah this is available maybe I can run it here uh what I'll try to do is just go back to my script remember I have done the change here and I'll just say instead.
Shueb Ahmed
Of running it on Google device full name I'll just change the parameter here and then run it on a different device this was on the 11 [Music] version and let me see if I.
Shueb Ahmed
Say control yes okay remember I have done the changes for the Blood Bank application too and then I'll just go here and.
Shueb Ahmed
Then trioral test so ideally in Real uh in um in real time practice the application should get failed here T execution should ideally get failed but with self feeling it will automatically here all.
Shueb Ahmed
Right so I just wait for the device to get B yeah the device has been B successfully I'll just wait for this to get started.
Shueb Ahmed
Yeah it's on [Music] going yeah yeah application got installed username put some wait times there password entered and then we go ahead and click on loging.
Shueb Ahmed
Right all right as you can see the test got passed without any failures even though I had change in my resource in my new build all.
Shueb Ahmed
Right so that's the uh beauty of uh uh the AI based self feeling which automatically understands that there is a change within my uh resource ID goes goes ahead and.
Shueb Ahmed
Then getes the uh the the locators and the elements from the back end so it's everything is been taken care at the at the platform level itself uh there is no prerequisite as such.
Shueb Ahmed
Right earlier we had a prerequisite wherein you would Pro you will be providing the you had to provide the capabilities for selfhealing all.
Shueb Ahmed
Right so there was one specific capability where you would be adding it and uh now that dependency is gone have moved to a new self feeling for the advanced version.
Shueb Ahmed
So where in this is taken care automatically so this helps in overcoming uh any changes uh within your applications if there are any new bills at the object property levels and this gets updated.
Shueb Ahmed
Automatically all right so arst got passed for the same script for two different builds all right so that's how the uh self feeling Works in in in real time practice all.
Shueb Ahmed
Right so now maybe I'll just quickly take you through the uh the analytic section which we discussed just go to the report section here let's.
Shueb Ahmed
Look at the reports in real time how the uh the reports can be helpful in understanding if there are any other failures uh.
Shueb Ahmed
Right so reports are very key uh to understand uh the changes or the uh know the failures within the application.
Shueb Ahmed
So it's just loading just a minute yeah um it's on I can just go ahead and then yeah so I have various different scripts.
Shueb Ahmed
Right different reps maybe I'll just take one show you how the report looks like yeah so this is the uh the Insight real time insights that you see.
Shueb Ahmed
So while the execution is going on you can can still get these reports right so for example while the test you are exing.
Shueb Ahmed
So the test can on for you know few hours or maybe more than five six hours if you are you know doing a nightly build or you know if there it's a long suit.
Shueb Ahmed
Of test so while the test is going on if you want to have the real time insights to understand what are all the failures how many number of test cases that are how many.
Shueb Ahmed
Number of test cases that are failed so that's something which is uh provided as part of the which gives details inside.
Shueb Ahmed
So you can further get into the device level metrics understand and see the status of your test okay how much time what is the duration and on which device you performed your test along.
Shueb Ahmed
With all the details like logs we capture all the uh logs the capabilities that you have used to pass and what are the uh result to it the desired result which which can be.
Shueb Ahmed
Very uh meaningful and uh useful for your analysis purpose then there are device locks we provide all the device metrix locks which can be helpful in case of there's a there are any crash.
Shueb Ahmed
That happens within the device while the execution is going on so this logs can be very handy followed by APM logs all the APM standard or logs gives you um you know insight about.
Shueb Ahmed
Your application uh level issues if there are any around APM server s so you'll be able to get it profiling more towards the performance metrics or uh providing the data in terms of the.
Shueb Ahmed
CPU memory and the battery Chan all right and then also the session video will be captured here so which you can plate instantly um for your test execution.
Shueb Ahmed
So that's something which you can monitor in real time and which can be further downloaded in case if you want to download this report.
Shueb Ahmed
So that's something uh you know from the report standpoint what we provide and then you can also share the link uh if you have to share this report to your various different stakeholders for.
Shueb Ahmed
Them to uh understand how the automation execution went ahead and then also see the failures and the analysis of it.
Shueb Ahmed
So that's more on the uh real time Insight reports of it and then further we have added a few more insights within my within our test.
Shueb Ahmed
So if I just go back to my recent say let me pick up yeah if I just um U if I have to get an overview on the different runs that I have done.
Shueb Ahmed
So I can just select it here select the test which you want and if you want to do a comparison you can just click on actions merge we do provide a comprehensive uh analytics.
Shueb Ahmed
In terms of providing py chart right which can help you understand how many bills that you had run and what are all the results how many number of cases that are passed failed there.
Shueb Ahmed
Are any uh incompleted ones you know so you can you get to see all these in the mer column you can.
Shueb Ahmed
Also compare the results all right for various different bills how many number of bills that were run and what is the pass percentage.
Shueb Ahmed
Right that you see if I look at it uh the uh L do has uh decent amount of uh uh the past ones and.
Shueb Ahmed
So with the progressive report the pr reports so that's the build comparison that we do so if you're running multiple builds build execution and if you want to merge it and get an insightful.
Shueb Ahmed
Report you can get this as well and this is something which can be shared among the team members so this is more towards on the uh uh on the uh the report the segment.
Shueb Ahmed
Side uh which we just spoke about during our presentations all right so that's pretty much it on the uh selfhealing piece uh we are ahead of I mean we are very much ahead of.
Shueb Ahmed
The time uh maybe we can quickly get into the questions if there are any questions I can go ahead and take it up and answer your questions all.
Shueb Ahmed
Right uh I can see a few questions coming in uh let me try to answer them so we have a question from one of our uh uh attendees uh his name is Arun he.
Shueb Ahmed
Says uh what what are the prerequisites to get started with the uh self feeling understand there was some prerequisites in your previous uh version.
Shueb Ahmed
Okay good question I think um I did cover this during my uh demonstration but how uh there are no prerequisites assets that are needed as we have moved to a new self feeling which.
Shueb Ahmed
Is 2.0 U you know there's absolutely no need for you to add any capabilities or any other prerequisits it's everything been taken care at the platform level itself.
Shueb Ahmed
So we have uh you know brought this feature within the at the platform level so you don't have to add any capabilities as such just go ahead add the cap required capabilities for the.
Shueb Ahmed
Integration standpoint and and select the device that you are looking at it and so the end point so everything will be taken care.
Shueb Ahmed
So you can go ahead and test and see with your different builds if there are any changes within your object properties uh test with different different builds.
Shueb Ahmed
So that you can see the differences all right so that can help you um from the maintenance standpoint yeah hope that answers your question AR all.
Shueb Ahmed
Right so there is uh one more questions from one more question from Priya says how maintainable are the test what percentage of how much percent can we reduce in terms of reducing the uh.
Shueb Ahmed
The from the percentage standpoint all right so if I understood your question so you're referring to the maintenance effort how much percentage can we consider to say that the maintenance effort has been reduced.
Shueb Ahmed
A good question person I don't have a definite answer to it so as we are um moving uh more towards the advanced versions of self feeling there are uh certain limitations at this uh.
Shueb Ahmed
Given point but we have tried to cover most of it uh but however uh in general if you take an average about 60% of your uh test maintenance would uh can be U you.
Shueb Ahmed
Know reduced at this given point as we speak right it can be and I don't as I said I don't have a definite number here but still it depends on uh the type of.
Shueb Ahmed
Of the application that you're testing and the type of issues that you have within your test and the type of changes that.
Shueb Ahmed
So about 60% you can consider so as we move on this percentage will um will will grow gradually as we move on.
Shueb Ahmed
Yeah hope I answer to your question all right so how accurately does the AI fix the scripts right so I mean um to an extent we are able to capture all the attributes and.
Shueb Ahmed
Uh values various different values as we have seen today in today's demonstration uh we were specific around the resource ID all.
Shueb Ahmed
Right so it can be any any attribute or any attributes right so we are not limited it can be uh the xath as.
Shueb Ahmed
Well your xath could have been changed if you have used xath or even the CSS path so the platform the feature is built in such a way that uh as this is driven through.
Shueb Ahmed
AI which can identify the change um within your build that you have executed and um you know it tries to fix the uh the elements it tries to update the element uh for your.
Shueb Ahmed
New build automatically so you don't have to worry about uh fixing the issues uh manually right so I cannot really give you a percentage or how accurate it is but.
Shueb Ahmed
Yeah for most of the applications uh we are able to uh the AI is is is is capable enough of fixing the issues uh at the script level all.
Shueb Ahmed
Right so that's something which um uh you know uh from the platform side where we have built this feature.
Shueb Ahmed
So which can help you overcome uh the issues that you are facing right now right so I I encourage you to go ahead and.
Shueb Ahmed
Then try try it from your side and see um how the how it helps in your scenarios could be a ones you may you might take some uh different applications and test it and.
Shueb Ahmed
Let us know your feedback on it yeah but this has been tested so you can be rest assured and U you know see the changes within your application site on the test execution hope.
Shueb Ahmed
That answers your question all right so there's one more question from SAR he says does the sell feeling apply for all of my script or will it work for multiple scripts it is for.
Shueb Ahmed
All of your scripts right um so when you say all of see basically you might have different set of uh test cases.
Shueb Ahmed
So irrespective of uh you know which test case or which script that you are running if you are executing it on your PE Lo platform it is in built with a self failing feature.
Shueb Ahmed
Which is which will be applicable for all all your all your tests you might have hundreds of tests that you are running it all.
Shueb Ahmed
Right and um yeah you so you can use the the self feeling uh capability for your different test and it it pretty much applies for all for all all your test scripts all.
Shueb Ahmed
Right that answers your question and then there we have one more question uh can we use self feeling for testing browsers and web testing as.
Shueb Ahmed
Well yes of course you can use it uh for for your browser based uh applications as well uh where this can be incorporated uh.
Shueb Ahmed
So this is something um where we have advanced for mobile and as well as the web application site so this is applicable for both mobile and as.
Shueb Ahmed
Well as web web browsers so there's one more followup question from Seer who says can I use a self feeling for my existing scripts yes of course whatever script that you have cre see.
Shueb Ahmed
It's not applicable for script that you are going to create right it is applicable for all your scripts IR respective of uh uh what scripts that you are using in uh.
Shueb Ahmed
So it is applicable as long as if you are executing on the ple FL platform the self feeling feature is inbuilt and uh this will be taken care I hope I have answered to.
Shueb Ahmed
Your questions all right uh so that's uh pretty much it from my side uh maybe if you guys have any other questions uh please feel free to reach us out I'll be happy to.
Shueb Ahmed
Answer them oneon-one and U yeah thank you so much for your uh time and the interactions that we had was uh really great I have see a lot of questions from different teams from.
Shueb Ahmed
Different people coming in so yeah thank you so much uh for your valuable time and hope uh the new selfhealing feature can bring in a lot of changes within your uh automation test automation.
Shueb Ahmed
For your mobile as well as web application so that's pretty much it from my side uh thank you so much again.
Shueb Ahmed
Yeah, have a good day, good night, good evening to you all.
Dinakar
All right, great, thank you, thank you so much for sharing your valuable insights Shueb, and uh, even as we draw this presentation.
Dinakar
To a close you've kind of been very practically sharing insights in the testing space so we want to thank you for sharing your wisdom with us and we.
Dinakar
Also want to thank each one of our participants who have joined in and set aside time to be here and uh learn about test flakiness test reliability and.
Dinakar
Also learn a bit about the selfhealing area that we're dealing that we're working on so great and I'm sure these few insights have sparked some questions sparked some ideas to think about to come.
Dinakar
Back so please feel free to reach out to us us uh you we've got our contact details we'll be shooting out an emailer with the recording and you can contact you can write back.
Dinakar
To us on that email itself and we'll respond to each one of you if you have any questions so thank you once again for joining in and uh have a great rest of the.
Dinakar
Evening or the rest of the day thank you so much everyone bye bye.
