TRANSKRYPCJA 🔗
speaker-0: What is your favorite automation framework to work with alongside Select itself?
speaker-1: Telling you miss my love but the way people today moving towards playwright and exploring new things. So yes, I'm also trying and adding support for playwright. But everything who will teach you is that that's your smart editor like IntelliJ or Eclipse when you write the code the moment you type dot and then all the methods appears like the moment you inspect any element. Nobody is there to tell a beginner that it is inside Shadow DOM which can help you directly integrate with your like any test management tool like G dot etc. So yes, that scale thing also we have considered, customize the test case template, script thing so all those things are there which has been thing from the art trainer or art blog or anything, any book can explain a few of the methods of that particular API. Wherever possible, like we used to keep trying to optimize like five, 10 seconds, how can we save in each and every step for our testers. One example, like in place of downloading the test cases or the, downloading these steps, we give the copy icon. just copy and paste it.
speaker-0: Welcome to Browser Stack Talks, conversations that turn testing pain points into playbooks. I'm your host, David, and today we're going to be talking to someone who's just done some phenomenal work and tries to demystify how kind of to do test automation. We have Sanjay. So Sanjay, I'd like you to introduce yourself because if I get anything wrong, ⁓ it's on me. But if you get it wrong, it's on you.
speaker-1: Thank you. Thank you so much, David. First of all, having me here in the show and ⁓ my name is Sanjay Kumar. I used to keep creating the productivity tools for the testers. It's been a decade, almost like now nine years I've been doing this and I'm enjoying this journey with the community support. We are building these tools today and ⁓ yeah, it's going and growing. That's a quite short intro.
speaker-0: Awesome. It's a great intro. so thank you for being here. ⁓ so as you said, you you've been working on loads of tools ⁓ for kind of everyone and you have done training and kind of you just demystify kind of how everything works. ⁓ because ⁓ as someone who's worked on the browser ⁓ and worked in test automation, there's like I don't think a lot of people truly understand how. complicated like a DOM is, right? They they just open their dev tools and go, ⁓ I I that that's the thing I want. That I'll that'll do. But that doesn't turn out to be the case. ⁓ so like I want to start with ⁓ how you've kind of got to that stage and then we'll build out more. So like ⁓ for anyone who is new to the automation space, what is the initial pain point or aha moment that kind of led to the point that you were like, I need to create selectors hub.
speaker-1: Yes, so initially like there have been always many many tools which used to solve this XPath and selector's problem but most of them were like always it's a auto generator tool like you use the tool you can generate the XPath or selector but none of them have ever solved the problem of like how someone who is new to automation word or just started how they can create their own how they can learn about the XPath and selector yes by reading the books, blogs or attending training, you can learn only those methods, what they will teach you. Like same any programming language. If you learn from someone, will just because a trainer or a blog or anything, any book can explain a of the methods of that particular API or language. But everything we will teach you is that that's your smart editor, like IntelliJ or Eclipse. When you write the code, the moment you type dot and then all the methods appears any error like exception handling or anything there it immediately highlight like okay there is a this error is there in your code so that was never there for selectors and x path like the moment you inspect any element nobody is there to tell a beginner that it is inside shadow DOM or if can be automated or not with through selenium or in automation so That was the thing. And the moment you write and the people used to write in control F in dev tools that are writing XPath. If they're making any mistake in the XPath and XPath is wrong, nobody is letting, telling them or nobody is there to tell them that this is syntax is wrong. So that's where I got the idea of this, that we should create something which will help people to learn about XPath selectors and DOM. The moment they inspect, they get to know, this is the thing. Like people used to spend hours of time just to figure out this element is inside shadow DOM. Wasting hell lot of time to find out like this element is inside iframe. So I mean like now things are different but ⁓ back in time there was a huge huge time which people used to waste just to figure out ⁓ this element was inside iframe that's why my script was failing. Then they go check in the DOM scroll up and down. So these were the initial things which pushed me. to create this tool. And also people used to always say that don't use an XPath and selectors tool because you will forget to how you will, how to write XPath selectors or how to learn about DOM. So that's where I thought like we should have something like an a smart editor, which will help people to learn about these concepts. And at least people will say, go and use that tool, selectors up, it will help you to learn about XPath. And that's what we were able to. Achieve that and people now like pretty much everywhere, every trainer and every Udemy course today had has the slacker sub in their course. So that's a good thing. Like with the help of community support and everyone, we were able to achieve that.
speaker-0: Yeah, I like the the key thing for me that you said there is that like ⁓ when people look at a a page and then they go, ⁓ I just need that thing, right? And like sometimes they they'll right click and DevTools
Load full transcript
will try give you something that makes sense, but sometimes it's incredibly way off. And I ⁓ while I'm a big fan of like the I the concept of Shadow DOM, ⁓ I've always been quite annoyed with the way it was designed because it was never thought with testing in mind. Not from the f the beginning, right? Like it was like, ⁓ this is cool abstraction. It's like, yeah, awesome. How do I test it? And they're like, Yeah, I don't know. Right, right. Like it was always that way. and it was a a conversation that I had with the ⁓ 'cause I do a bit of ⁓ spec work with the W three C and as I had that conversation with the people who wrote Shadow DOM and they're like, Well, you know, you'll figure it out. I think
speaker-1: Yeah, it was very, very challenging, always, Serodong.
speaker-0: Yeah. ⁓ so like you said like you've got tools that allow you to kind of you know you've spotted this mistake, you do this. ⁓ and I think kind of a lot of tooling nowadays is trying to kind of get into a space, especially in the AR space, but we'll get more into AI later. But like how you know, a lot of tools are like going, here's this thing, this auto generated thing ⁓ that you need. ⁓ How do you feel with kind of like the work you've done that people are moving more to this auto generation without necessarily understanding the underlying how the DOM is generated problem?
speaker-1: ⁓ yes. So auto-generation like always been there even, mean, like, ⁓ earlier we used to have a lot of play and record plugin, which used to like generate the script, but, ⁓ for the demo I have like experienced also. And I'm sure like you must have, ⁓ you will also agree on this, that we, for this demo purpose, most of the time, these auto-generation of the script, re-running the script always works. Real pain comes when people used to generate the script for the enterprise application or for those applications which used to rapidly change and where like something called auto healing or those kinds of stuff is required where locators are changing, pages are changing rapidly. So, and sometimes you, you know, the business logic, like what kind of test cases should be automated for my application, telling you to the tool or a tool can understand automatically. Not every tool or every solution, even with the AI today has achieved that thing. Yes. In demos and all I have seen like, ⁓ wow. This isn't like people used to give it that login page or register page. But the moment it used to come with for the maintenance, ⁓ for the big enterprise application, those are very, very complex. ⁓ is like not auto generated script for automation have been working there. And this is one of the region that, tools like selenium playwright and everything has been always a huge success. The percentage of people that were using auto generated one and the percentage of people there's no competition. Like I don't think like at all there's a competition. Everybody this is the one of the region. Like everyone used to choose
their to write their own script rather than going for because maintenance located generation and all putting your own business logic and maintaining with the tool. The cost is also one of the matters. there so yeah.
speaker-0: Yeah, I I think that's true. And it like i i it's one of those ones for me, like having worked on Selenium for a while, like ⁓ because of we knew how many people needed to write their own things from scratch, ⁓ our opinion was that like and some of it's right, some of it's wrong with Selenium is that like you need to be on control with the whatever you're doing, right? And from there you can kind of build out like when you should wait, how you should wait, when like what locators, like and and navigate down because you know it better ⁓ better than some that like things move around and then you get all these flaky locators then people go, ⁓ my tests are flaky. ⁓ and it's ⁓ it has that like l follow on. so like how does ⁓ Like for people who don't know, how how does like Selector Hub kind of help them with those complexities to kind of understand and then go, ⁓ they then themselves have this aha moment with kind of how everything works?
speaker-1: Yes. like on the pre continuing from the previous question, like even these companies, those who used to auto generate the scripts or the tools, those who even their clients and their users used to use selectors have in those situations where there is their tool or because like you mentioned that their tool didn't understand sometimes it used to give the flaky locators or like very long from Taking the starting from HTML DOM from top to that particular element. And if anything breaks in between it goes and they keep trying like backup locators, backup locators, something like that. And that used to kill a lot of time. Now this thing that, ⁓ how selectors hub helps in such kind of situation or how selectors hub used to generate the best part about selectors hub is that like when you auto generate the selectors or X path, it used to look for that particular element. and try to look for the stable element nearby with this parent. It's never generated from the top. It always goes to generate for the stable locator. And then you have the option, you have the full control over there because you will be seeing that like, okay, this is what generated. If you feel that this could be dynamic, it will suggest to you that this could be dynamic. Then you can ⁓ change that. There are a lot of customizations available like with what kind of attribute you can generate. And also you can generate some with respect to some relative element as well. So it always goes for that very stable one. And because you, before putting it in your script, you have, like you mentioned, you have the control there and you know that what I have put tomorrow when it will break, you know that because of whether it break because of locator or whether it has been like any actual issue in your application or a script something. So that's the thing.
speaker-0: Okay. That yeah, that's awesome. ⁓ and I think kind of even if you're not using Select Hub, I think you've given people a good bit of nugget there where you're kind of saying like find the st most stable thing and then you can always build off that, right? Like ⁓ the amount of times ⁓ I've seen people go, I've written this X path and you look and it's like it's literally from the top of the HTML document all the way down to that element. And you're like no Please don't. Like that that's scary. ⁓ but you know, ⁓ and again, I think there's ⁓ like it's always been an education problem because ⁓ the DOM a lot of people think is like, ⁓ it's not that complex. ⁓ and I think kind of ⁓ especially as we build out more and more modern applications, ⁓ I think that complexity is just growing. ⁓ and like I like to joke it's not really a joke, but I like to joke with people that ⁓ the browser's kind of an operating system these days. ⁓ and it's like you gotta think of it that way. ⁓ and I think ⁓ for mobile, I think it becomes a really important if you're doing like native applications because of how you think about things. I think that becomes ⁓ more acute. ⁓ is that what you're also seeing kind of with like people you're working with ⁓ in that?
speaker-1: ⁓ yes, it's like, there is also a lot of challenge on mobile site, but we have never gone, I mean, so far, we still not, ⁓ added support for this mobile native applications. So we are still not there. we are, mean, like, because I have been the expert or maybe like I have been this side on browser site. So have been developing more and more tools for more and more problems rather than focusing on some area, which was not my expertise. So. haven't touched that that much tried multiple time, even still trying today to solve that problem. But because I have got so many ideas in this side that okay, let's focus this there's a so so many people whose problems could be solved with multiple solution with multiple ideas. And that's where we have been adding more and more tools, XPath and selector. This has been like quite an old thing now. Yeah, focusing more, more and more things on the other side.
speaker-0: When people start running things at scale, because I think that's the ultimate goal for people is to be able to do these things. and you've already told people, like, when you're focusing on these things, focus on stability from the start. ⁓ but you can tell people that and that's not necessarily how they act with it, right? ⁓ so if we kind of how like what advice would you give to people to kind of give them granular control, but also at the same time go Keep like building out for scale. and ⁓ like, you know, even if it's a case of people moving ⁓ who are moving from low code or kind of no code automation tools, like how do you kind of help people go through that process of like changing the way of thinking, especially like finding elements, interacting with
elements, writing tests as a whole to be able to scale that out?
speaker-1: Yes, to army like in place of the way earlier we used to do like one one element and because today like everybody like you said from look low code or completely no code they are going so in place of just blindly going there not having any control better to be in the sink of like the way you were doing start using those kind of tools which can speed up your process. But at the same time, should be like, an eye on these locators and element side because in when you write the automation script or generate the automation script in every line of your code, your framework setup that is always constant. Every line of code that something like driver dot find element or any framework you're using custom framework dot click dot send keys, that method is always going to constant. The only thing which used to change in each and every line is that the locators, because that is the most important thing in the automation. So that's where like you should spend some time, but at a scale you can use like tools like selector sub or anything which can give you, or you can get your complete page locators in one shot, but use that's where you give time. Rest of the time, rest of the things you can boost and make it fast.
speaker-0: Amazing. Cool. ⁓ so the the other thing that I kind of wanted to go to is that ⁓ you've been doing some work ⁓ in accelerating QA workflows ⁓ at the moment. ⁓ trying to kind of change how people think to the way they execute. ⁓ or at least that's the way I've when I've been reading some of the things that you've been doing. ⁓ so Before we kind of die deep dive in, ⁓ like one of the things that you've also been working on is ⁓ like a test case studio. Do you wanna just explain to people what that is and ⁓ how they use it?
speaker-1: Yes, so the idea was very simple that I have been seeing people that like ⁓ people used to generate the X path and selectors from this tool. But then I thought like, can we have something which can help people to get the steps and screenshot also automatically and not the steps like any other recorder plugin used to generate because most of the recorder plugin what they used to do like they used to generate the language, kind of, which is a kind of like, looks like a machine language and a kind of like automation is script. I was thinking that can we have something which will give a plain English sentence, which we used to type in a bug report or in any test management tool that this is how you reproduce the bug for a developer. So that was the idea for this tool, which make it a big, big hit out there. And ⁓ so it basically used to generate very plain English sentence for your each and every action. And when you are doing your testing, you just open this just by clicking on the logo. It will capture each and every step, what you will perform in a very plain English sentence. You can directly copy them, paste it to your developer or save
it as a documentation. And then also it captures X-Pars selectors, a screenshot and lot more things. And in fact, like it also helped like to quickly write your automation script very fast. So that was the idea behind this tool to get the steps and screenshot. And without any like, ⁓ there was a Selenium IDE before, which was widely being used, but, ⁓ like a lot of things were there, which could be improved, ⁓ in that tool. people, ⁓ when I started my career back, I used to use that, like the first training I got before Selenium was Selenium IDE. that this is how you used to record. So that time I was like ⁓ amazed. this can be automatically done. Just click and it started back in like after three, four years when I started developing tools, that's where I thought like in that tool, like first I used to open, then I have to type project name, then test case name and lot of things I have to do before it get started. And ⁓ that time I didn't get this idea that I can ⁓ like add a execute script functionality as well. But there I thought like a very minimal thing that if I have to record steps, there should be something click and just go. So, ⁓ so that was the simple idea, like in almost like 10 plus tools today, my every tool is like one click. So I have always be in that thing that we are going to save a very, very minimal time for our users, for our community, make it as simple as possible. that they shouldn't be doing anything. Some tools you don't even have to use. They will automatically be used. Some of the tools if required only should be, there should be a click. So Testcase Studio was developed with this concept that you just one click and you get everything. And that's the win for this. This is like today, top rated recorder plugin. And now we also added that execute script. So record and play because Selenium ID is not there. So this is there people are using. ⁓ this tool to replay the script as well and it quite quite a stable one.
speaker-0: Yeah. Awesome. I think like I think it's I think it's important ⁓ to have tools like this. I've always been a big advocate for things like Selenium IDE in the past. it's a it's a really poor term and I can't think of a better term, but kind of I always saw it as like a gateway drug, right, to get into test automation, right? You do you could get in there and just like go, ⁓ I can I'm I've done stuff, right? And feel confident in in your ability to do it and and do it well. ⁓ it didn't teach the the the the downside is it didn't teach you kind of the the problems that I think kind of you've then gone on to solve with like Selectors Hub, right? You were like, ⁓ yeah, this this works and it runs. ⁓ and it becomes tricky, I think, is the best way. And then but you don't s learn any of those those good sides, but you kind of It does make you feel good and it makes you feel like you're being productive. And I think that's awesome. ⁓ and I think like for me, it's important to have meaningful bridges between kind of exploratory testers, business analysts or product managers, and kind of automation engineers. ⁓ and I I think that like to me, that's what your tool is trying to do, is like be that good bridge between people with the kind of plain English side of it, isn't it?
speaker-1: Absolutely. Yes. Yeah. So that was the whole idea. Like just keep it as simple so that like people immediately get that. Like you don't have a generally like we used to write open this website and go here then click on this then type on this thing for developers. You just get automate automated thing handy copy and paste. That's it. Nothing. So in one go because sometimes it happened like after 50 steps you got a bug. And it's a pain point for the tester to recall all the steps, hit and try to reproduce that and then typing all those steps manually. It's a big time pain.
speaker-0: Yeah. No, I can believe that. ⁓ and like that that's the downside that I was talking about is that like if people don't understand it then it becomes other other problems. ⁓ the tool also kind of tries to help people understand ⁓ kind of entire pipelines ⁓ for when it comes to testing. ⁓ kind of it has integration points that pe people can do it and then they can kind of ⁓ run their automation ⁓ suites at scale like ⁓ you know can run hundreds of parallel parallel builds on t something like browser stack as an example. ⁓ how does kind of having that ⁓ precise data from kind of locators, recorded steps, how do you kind of translate that scale of information back into people and go Where like your initial goals one click. Now it was like, how do you get that one click of understanding ⁓ where things go wrong?
speaker-1: Yes. So recently we added that feature where you can get this complete thing because for locators of each element on which you perform the action, we integrated slackless sub in test cases studio where it used to generate the locators for each and every step. And now we have also added the feature of like, can get the script as well that complete script, executable script, which you can execute. and export as well. in that like everything is there for you that all the stable locators and everything which you can download and you can scale that particular test case and you got everything in one shot and there you used to get a lot more information not just the steps you get the your system information, browser information, whichever is required for your particular test cases, documentation or the bug filing. or in fact, reproducing the bug automatically, that is also there. it also has a feature where can automatically, like kind of pseudo code you can generate for ⁓ whichever framework you are using, irrespective of Selenium, Playwright, WebDriver, whatever you are using. There is a feature that you can generate, append on that particular locators, that particular command framework, and... When you export, can download that just you have to add dot click dot send keys,
whatever method you use, just add them and your script will be generated and that will be very, stable. So that's how you can scale as well. And not to mention, but yes, we do have the like integration of like these kinds of test management tool where you can directly upload or file the bug in your test management tool with the pro version of this that which can help you directly integrate with your. like any test management tool like G dot S. So yes, that scale thing also we have considered and a lot more things like you can ⁓ customize the test case template scripting. So all those things are there, which has been thing from the productivity perspective. So yes, those things have, we have been thinking like wherever possible, like we used to keep trying to optimize like five, 10 seconds. How can we save in each and every step for our testers? In place of like one example, like in place of downloading the test cases or the, sorry, downloading these steps, we give the copy icon. just copy and paste it to your developer direct from the check. How can we save that minimal, minimal time that was like really required for us?
speaker-0: Okay. That sounds really cool. ⁓ and so kind of that leads to kind of trying to improve ⁓ speed to resolution, ⁓ as I like to call it, or most people call it but like ⁓ is every ⁓ with every deployment. ⁓ I think it would be a miss for me not to ask this question, especially because like AI is everywhere. especially around the space of kind of fixing broken locators automatically. at runtime. ⁓ what is your take on self healing locators versus ⁓ trying to get people to get it right at the first time?
speaker-1: yes, I always, never been on that side of a hundred percent believing in the self-healing things and auto-fixing. I have seen like tools which have been like, keep doing for this, this thing, but it always required a lot of cost. And, ⁓ then you have to share your code and everything with them. And this would be having the previous image of your dorm and then compared to that. not in that side, I believe like the day it will be achieved a hundred percent. like auto healing or self healing this automation things because everything is running all around this auto healing like this web automation selenium playwright or all the tools this everything is running around this only and this is where we all are writing by ourselves. So it's still not that I think like it is a hundred percent achievable or not yet achieved that hundred percent. So mostly I believe in selectors healing or expert healing. where you can verify your all locators like slactor sub has this feature that where you can paste your complete script go to that page it will get all your locators from your page and verify and shows them which all got changed and then you change it by yourself on that i believe because from like we earlier we discussing that we will be having the control that what has changed in the script and where we have changed because sometimes a script automatically, it will be done, but then you will not be sure that which locator it has changed or if there are multiple elements, is it directly pointing to that particular element? So that doubt, I mean, like it's still, even if AI is there, always, I'm not sure. if everybody or people having that confidence a hundred percent that when it has been fixed by this thing, will it be always pointing to the right element or it could be that. to some other element. And if anything broke in production, can you say that 100 % it is the script failed or the my application failed? So that confidence, I don't feel like it's still there.
speaker-0: Yeah. I I'm I'm in the same boat as you and kind of it was a questi it's a similar question that I asked ⁓ Benjamin Bischoff in a previous episode. ⁓ we were talking about the work he was doing at Travago. And ⁓ like we like for me and and him, and it sounds like you too, like the the key thing is like determinism, right? You want to be able to know that you understand why something is that way in the first place rather than just kind of believing it. ⁓ and I think kind of it's ⁓ it's a very interesting space, I think is the way to probably think of it as like how all of that can change all of a sudden. ⁓ and I think kind of ⁓ for me I always want to like if a test fails, I wanna know why. and be and like but I and I want it to be failing f as fast as possible so that I can do quicker feedback loops. So ⁓ that like you know. We're living in an age where you can easily get distracted by your phone, right? Like how can we stop that?
speaker-1: Yes, yes, I will like very well mentioned that when a test fail, we should be having this confidence that test fail, not our script fail, not our like application or something. So it should be when a bug is there, I script found it should always, we should always be in a confidence that we know that this is a test fail. It shouldn't be a doubt like something got broken. Then we are in a doubt is our script broke or the test broke. So if you are in a doubt, then the complete purpose of having this script gone. So don't have a script at all. if you have the confidence, should always be there. OK, if the test break, the test break.
speaker-0: Yeah. Yeah. Yeah. And I think kind of ⁓ like we've discussed it earlier, like ⁓ there's multiple things that can cause a test to fail, which is like locators, but it's also n ⁓ knowledge gaps. ⁓ and a knowledge gap to me is where flaky tests suddenly come in, ⁓ where like you've made this assumption that something's gonna be there in time, you try to interact with it at that time, it's not there yet. It does. ⁓ and ⁓ so I think kind of the the key thing and we're both agreeing is that a key a good solid foundation beats every bandage along the way. ⁓ so for teams aiming to achieve true continuous integration, where do you see the most overlooked ⁓ productivity bottle at neck right now and like how are you aiming
to solve it?
speaker-1: ⁓ Can you explain like little bit more your question?
speaker-0: So ⁓ like where like if people are doing things and they've kind of got like test case studio running and they go This is running really well ⁓ but whenever I do it I get stuck here or it feels sluggish or ⁓ you know it takes a long time to get feedback or ⁓ my tests are flaky. ⁓ like those can be productivity bottlenecks where they're kind of really stuck waiting for something to happen. ⁓ what is the like next one that you're trying to aim to solve ⁓ for people?
speaker-1: Yes, good. So couple of things we have also solved recently because ⁓ like with this, these automated things, the new new tools every day out there and with the AI, everybody wants like everything to be done on the fingers and should be very, very quick. And now the development and everything has a lot been easier. So yes, we have solved a couple of bottlenecks for our testers recently. I mean, like it's been quite some months now that, ⁓ thought like, ⁓ earlier you mentioned in something exploratory tester. So we developed a tool where like we thought like, ⁓ the moment every release or any build we used to get. So there is always a set of test cases, which every testers always do like a very basic testing. all do like checking every input box is enabled or not, not disabled checking. There is no API failure call is there. So every time you get a bill, we all do those basic check. Like every test does that. So we thought like, these are the set kind of things. Why don't we automate them just in one click? Everything got done for you. Then accessibility testing. There are some defined standard like in your web application, if you are going to test the accessibility, these are the basic standard which need to be followed. So why don't we automate them? So these basic checks that your page should be having that we have automated with this exploratory testers and it immediately like adopted by the community and so many people are using today. Just in one click, it does your all like basic testing for everything like accessibility scenarios and hundred plus test cases just in one or two seconds got covered and then you can focus on your business use case. So this is one bottleneck was for the testers. And now they are, that has been solved. course, like you will have to do your business use cases and other testing, but yes, we are, we have done this thing. Then other major bottleneck for our testers and developers as well. That for all our testers, the greatest weapon is a screenshot. The moment we used to get a bug or something, first thing we do, we capture immediately the screenshot. So I have seen there was a big problem bottleneck for our testers. in the screenshot taking, we thought like, okay, a screenshot, it is a very simple thing. can command shift R and then sorry, command shift and four in Mac and short, shortcut that take the screenshot. But then one day I realized that there's a
problem. The moment we add that screenshot in a Jira or service or developer, he always asks, most of the time they ask which page it watch. Like, can you give me the URL for that page where you have taken this screenshot? So I thought like, don't we have something that in a one click, take the screenshot with URL. Right. So in your screenshot, there will be a URL. just click automatically, the screenshot got captured and it should not get manually. You can take with the URL, but the problem was it used to capture your personal information, what you have bookmarked, what extensions you have installed. So there's something which captured your screenshot, the area which you want. have the URL without any personal information. So with that basic idea, I developed this tool, Screenshot with URL, which is helping many, testers today and just very short time it has solved. like these kind of basic things, then there was a tool like we solved another great problem, which is in fact like ⁓ one of the most used tool today for us, Check My Links, which basically checks finds the broken links for the ⁓ testers on any webpage with just one click. So you see all these things are like one click and that used to quickly verify like earlier we used to write the Selenium script and even today many people write double forward slash a get all the links then HRF attribute getting the HTTP status code. This tool just in one click you get like entire page got a scan. And get those. there are like these kind of basic, basic bottlenecks have been solved for testers in many ways. And we are trying to solve with this so that like, ⁓ even with the like AI and everything, whatever you're developing and doing, things will help you to speed up your basic tasks, which you will be not doing through your AI tools and everything. ⁓
speaker-0: So, ⁓ I think kind of ⁓ like one of the key things that you're trying to do is ⁓ reclaim wasted time with people, right? So allow them to be more focused on getting stuff done. ⁓ and I I think that's awesome. I think kind of it's ⁓ everyone's working ⁓ so flat out at the moment, ⁓ in every role, right? ⁓ and I think kind of AI is ⁓ making it both easier and harder at the same time. It's this weird paradox that I'm seeing, right? Like it's everyone's j generating way more code and features. ⁓ CI systems are just struggling to keep up with it. I saw someone y ⁓ yesterday posting on Reddit the their their wife had been vibe coding and in about half an hour generated seventy five PRs on their on her on their website. ⁓ and it creates interesting problems. So ⁓ I'm I'm glad someone's trying to look at it ⁓ in a meaningful way. Awesome. So ⁓ Sanjay we are about to head into our final section. ⁓ this is a quick fire round. ⁓ the idea here is that ⁓ I ask you a question, you have about 30 to 45 seconds to give an answer. ⁓ so the main thing is mostly because I like to waffle. ⁓ this is to try to keep everything short. And I can't I can ask follow up
questions, but I really shouldn't. ⁓ but it's also kind of fun questions. ⁓ are you ready?
speaker-1: going yes I will try my best
speaker-0: Awesome. ⁓ What is your favorite automation framework to work with alongside Select itself?
speaker-1: Selenium is my love, but the way people today moving towards playwright and exploring new things. So yes, I'm also trying and adding support for playwright.
speaker-0: what is one book or resource that every QA lead should be reading this year?
speaker-1: Okay. So I'm not ⁓ a book reader ⁓ much, but yes. So I will pass this question. I hardly remember the book names because I mean, like after college, a very few books, some of like one Pradeep Sunderajan has written that I started, but then like never completed much of the books. yeah.
speaker-0: That's cool. ⁓ we'll edit that out. what do you think is the most underhyped QA metric ⁓ teams should be tracking? If you weren't building world class testing tools, ⁓ what career would you be pursuing?
speaker-1: I'm loving this entrepreneur things and the one which will not be controlled by ⁓ someone and like a kind of freedom. And that's one of the reason that we never gone for funding or something because I have always believed that if I will build something, I want to build something that can be used by everyone. So this is where like for me, it was always like more than making money. like I have been always in that side that if something has been innovated. very good that should reach to everyone and everyone should be able to use that. Otherwise, like if we will make paid, yes, you will, ⁓ we'll earn the money because that product and idea will have the potential to get money, very few, like maybe one or 2 % people will be using your innovation and your ideas. But if you, ⁓ I really wanted to be used by everyone. So yeah, on that side that I wanted to become someone who can build the products, which can be used by everyone.
speaker-0: I'm not gonna lie, that is the exact same reason why I've stayed in the open source space for as long as I have. ⁓ it's ⁓ I grew up in the the age of ⁓ like Linux coming out and people going, you know, like this is about making everything easier for the world to be able to access everything. ⁓ and it's like from there I s I did a bit of Linux and then kind of moved into the browser space because I thought the web is this awesome thing too that we should be sharing. It's got his problems, don't get me wrong. but it it was it was part it was the main reason they kept me at Mozilla for as long as I I was there is that ⁓ like the web is this beautiful thing to me. and open sources is another beautiful thing and we as humans should be sharing as much as possible because together we kind of become better. as a world. So awesome. ⁓ why does the world still need human ingenuity in an increasingly AI driven s landscape?
speaker-1: Thank Definitely one, one of the things is like maintenance and ⁓ sometimes like when user will ask and user will come to you, like you should be having the answer immediately for that. I mean, like in many, many aspects, there are like many things that like many times if we are like, ⁓ we also using, if I take the example of my situation right now, we also use the AI for like in some cases to develop our tools as well. Now today, pseudo code or a piece of code. But we are keeping the complete control of like what we are developing, not everything with the AI, but we should be knowing because when user comes today or tomorrow, sometimes Claude is down. Sometimes we don't know what AI has developed for us. Then what we will answer to our users immediately. Many times I have seen that ⁓ we have to release a bug immediately. was a Claude was down for one day a few weeks back. we generate everything through AI, then like who will maintain. I believe like it's really required. I mean, like I'm just giving the short answer. There are many, use cases. There are many scenarios we can discuss on that where humans are really required. One of the like real example is this, that we should be knowing that what we are developing, what we are having, and we should be having the knowledge. What if like today they're giving everything for free. It's not going to be remain free. Many companies already started realizing that, the human was much, cheaper than what the AI is today. I believe that like everybody, I mean, like, just really, really wanted to mention this thing here that people think like with the AI development has gone much faster and it has become very, very easy. But I believe development was never a big challenge. You could always hire two, three developers and get things done, get the product developed. Yeah. It has made things faster, but the real challenge come for the products or anything or tools like this is the user. adaptability, maintenance, and letting make your people use your tools. Will that AI can do each and everything that future? mean, like we are still, we all are like, still looking for that answer. We all are discussing all the conferences and everywhere people are talking, talking, talking, but we'll see the real picture in coming years. I feel like it's still everybody is on that stage where like, we, you and all are like in confusion in dilemma. Will that completely work or not? Or it is, it is a bubble or what? So let's see in the future.
speaker-0: Yeah, though I a hundred percent agree with that. ⁓ Sanjay, this has been really ⁓ incredibly insightful. ⁓ I really appreciate your time today ⁓ and your relentless focus on trying to make people ⁓ their lives easier and ⁓ thank you for sharing your expertise. ⁓ if people wanted to follow what you're doing, where's the best place to see what you're doing?
speaker-1: Our best place they can follow me on LinkedIn and they can have our tool testing daily, which gives all the updates from me and from the entire community. So that's one of the best thing that people use today to get the latest testing updates from me or from the entire testing community.
speaker-0: Awesome. ⁓ so to our listeners and people watching us on YouTube, ⁓ if you want to kind of follow all of ⁓ that, ⁓ we'll make sure that it is in our show notes. thank you all for ⁓ listening and please make sure that if you haven't subscribed or followed that you do that on whatever platform you're doing and we'll we'll see you in the next episode. Thank you all.
speaker-1: Thank you so much, David, and thanks everyone. ⁓