TestSprite AI - Basically Copilot for Test Automation! (AI Agent Powered)

0:00 Everybody is talking about how we can make use of large language models and AI and AI agents to do code development and application development. But nobody is really talking about how we can make use of the LLMs or AI for doing the software testing. This tool the test sprite can help you with the power of the AI agent to do the software testing in much much easier fashion. They can do a lot of different testing and that's exactly what we are going to be discussing in this particular video. So let's jump right into it. So we have already discussed about the AI agents. If you have not watched already in our exotition YouTube channel, we talked about what exactly is the AI agent. So AI agent is an fully autonomous system that operates independently over extended period using various tools to accomplish the complex task. It can

0:46 perceive, desire and act on an environment based on the given goals. That's what is the AI agent. We also talked about what is the agentic multi- aent system which is pretty much like the group of agent which are going to be communicating to accomplish a task. That is what this multi- aent system really does. And they can communicate, they can delegate the task and they can coordinate the task among the agents to achieve a specific goal. And while we have all this information in place, we are now going to see how the AI agent in the software testings can be done. And this can be done with the power of this tool. The test sprite the test sprite is an agentic multi- aent AI test agents or a collection of test agent which is going to do these operations for us. Well the test sprite AI agent can

1:33 perform a lot of different operation with the power of these agents by doing test object inspections generate the test plans for us create the test cases execute the test cases and also analyze the test results to see how exactly is the test execution happen. So you can see that all of these operations are going to be done by the army of the AI agents which are going to be sitting in the test pride uh cloud which is going to do all these operations for you and that is the real power of this uh whole system that we are going to be discussing today. We'll see a complete demo as well and you'll understand how powerful they are. And the next question that you naturally get is like what kind of test that the test sprite can actually do. Well, you can think of anything that you wanted to do the testing of your application. You can do the API testing. You can do the UI

2:18 testing. You can do data testing. You can also use the test sprite to do AI agent and model testing. So all of these operations can be achieved with the power of test sprite which is which is quite amazing. And also there are some more awesome features that you can think of in the test sprite is you can run the test in a scheduled mode. So no more manual trigger needed. You just have to schedule it and the test execution is going to happen. And also you can get the PDF report once the test execution is done even in the email if you really wanted to or you may you can just download it from the download button over there. The email PDF report or the PDF report in general is going to be so descriptive that it's going to have the entire information of how the desecution happened what are the failure what are the reason for the failure and how you

3:05 can even fix those issues. So all these informations are going to be there in that particular PDF file which is amazing. And then you can also see the failure logs with the test execution details along with the trays and also what is the solution to fix this particular problem within the test sprite itself. So it's pretty much like a testing expert who is going to be sitting on the on the cloud with all the power of this AI agent and doing all these operations for you. And most importantly you don't even have to write a single line of code. Everything is going to be achieved with the power of test sprite for you. And there are some general question that we mostly get while we use these kind of tool is that can we use this test sprite to run in the CI/CD pipeline. The answer is yes. You can export these test code which is

3:51 generated by this test sprite and then you can even check in the code in the GitHub directly which is even more straightforward. And the next question is like can the test sprite run uh the local application testing? Of course, yes. Even though it runs on the cloud, you can still test your local applications using the tunneling option which is also available. And is this human in loop while the tests are being executed? Of course, yes, you are still in loop. You can use the agent chat and also you can control the test execution. You can also add more test cases if you really wanted to or if you want to enhance the testing. You can do all of these operation using this particular test sprite. So you still are in loop for all of these operation. not the AI is going to take over all the job and get the job done for you. And how about the security because it is going to be running on the cloud of the test. How

4:38 exactly it's going to be maintained? Well, it is sock 2 uh certification compliant which means it is highly secured and I think they are still in the process of getting more and more. So that's all the general questions being answered which is more common question that I always get while we use this new tools. This is already been in place. Well, as that said, we are going to see a quick demonstration of how this test sprite is going to work and how you can get up and running with the community edition of the test sprite which has already given all the features that you can use in a pro level and I'm going to show you that in this particular demonstration. All right. So now I'm in the test sprite page over here and you can see that this is the homepage and they have got all the information like how the test sprite actually works and what are the capabilities that the test sprite has and what are things that you can do with the test sprite over here. Those are the information that you have

5:25 got over here and I'm going to quickly show you how you can get started. All you have to do is just click the sign in button and then you can sign up uh for the free edition. I'm also using the uh the the starter plan over here as you can see. Thanks for sponsoring this particular plan for me. Testbrite they have given this to me but you can use this plan for $119 per month and you can start using it and you can see that the model they have got over here. Right. I'm going to quickly show you how we can write a very very super simple API test in this particular demonstration and we talk about the UI testing in our next demonstration but today in this demonstration I'm just going to focus on the the UI testing the API testing to see how it can be done and the application that we are going to be testing this time is going to be the um the EA app uh API and the application that we are going to be testing in this

6:11 particular uh time is going to be the EA API.somi.com somi.com. So you just have to go to this particular link and you can see that this particular API is actually uh holding the authentication for you uh and then you need to uh do uh the authentication login first and then you have to perform the rest of the control operations over here something like this. So you need to call this component uh by doing try out for for instance if you just go to the uh products over here and if you just go and do a try out you will see that you're going to get an unauthorized exception. So first of all you need to authorize uh you need to get the beer tokens and then use the beer token to do the authentication of these APIs. So you have to tell these information uh to the test sprite while you actually uh start

6:56 doing the testing and all the test cases are going to be generated for you by the test sprite with all the power of the agents that it has got and then it will do things for you. It will write the code, it will run the test and also it will generate the reports everything for you. You don't even have to do even a single thing. You just have to import this uh this particular API and things are going to be done for you. So I'm going to go grab this API link over here. Uh and I'm going to go and I'm going to create a new test. So I'm going to say uh UI testing of EA app. Something like that. And I'm going to hit next which is going to uh bring me up to a location where I can add the backend test. You see that now this is the backend test. You can also add the UI test by skipping the backend test

7:41 over here. I'm just going to focus on the uh the backend test over here. So I'm going to say EA API tests and this is the URL and the authentication is going to be uh beer token. I'm not going to pass the key directly because that's going to be a fly in this uh in the security system. So I'm not going to do that. Rather I'm going to uh ask some questions over here. Um this is an optional information that you can give because it'll automatically discover things for you. But I'm going to give those optional information uh so that it can be helpful for my tests to be generated in a more robust fashion that I'm looking for. So I'm going to say uh can you read the swagger documentation and then understand how the APIs are designed in the schema. Now you should understand how to authenticate the API

8:27 and these APIs are authenticated using the authenticate uh URL which is this one this endpoint. uh and then this is the username and password that you have to pass in and based on the beer token you can authenticate all these APIs and then perform the positive cases, negative cases, edge cases, security testing, data testing, boundary value analysis, whatever that you throw it, you can just keep on adding all those information over there. Those test cases are going to be generated for you. Look at that. That is the only thing that you have to that's a prompt engineering to be honest. You just have to give all this information and things are going to happen for you. And once this is there, I'm also going to uh include the swagger documentation. So you see that this is the swagger documentation of this particular API. This will help the AI

9:13 agents bit more to give a context like how the uh how this particular schema is going to look like while it is testing them. So I'm just going to copy it and I'm going to create a PDF file or doc file or MD file whatever file that I wanted and then I can upload this file so that this information can be helpful for the AI agents while doing the uh test execution and also during the test case generation. So that's what I'm going to be doing over here. So I have copied that everything in the PDF file. I'm just going to open that and I'm going to upload that file over here which is going to be this one. I have just uploaded that particular file. I'm not showing my file system there. That's the reason why I'm not showing it. see that this is the swagger. PDF file. I have just uploaded that particular file over there and then I'm going to uh hit the next button. So basically I'm just

10:00 going to do the backend testing this time. No UI test. And I'm going to hit next over here. The moment I hit the next you see that now the AI agents are in action. They are going to read the swagger documentation for you, the APIs for you, grab all the information from where it has to do the testing, how to do the uh authentication and it generates all the test cases for you. Look at that. It has entered like 20 test cases for authentication testing, product creation testing, product retrieval testing and there is the security testing. Uh and once I hit next over here, you see that it's going to do all of these operation for me with a level of the priority that it has to take for these tests. Uh and what these tests are going to be doing over here,

10:45 something like that. And if you think that it is missing some of the test cases, you can still ask the AI agents to do it. Let's say if I'm going to say uh can you create um maybe create a component as well. So I'm going to say create uh component uh successfully. Uh and I'm just going to grab these information directly and I'm going to paste that over here. So I'm going to say test the API providing all the required field to create a component uh component ensuring that it returns the success response with correct product details. So I'm going to also tell that information over there. And if I'm not really happy, I can still keep on adding those information. I can set the priority high and then I can say uh can you do a boundary value analysis

11:31 testing and that that will be added for me as well. But you can keep adding more and more tests if you think these test cases which is suggested by the AI agent is not enough for you. And once we have everything over here I'm going to hit this next button and you will notice that the test cases are going to be generated. Uh which means like coding is going to be written for you. is going to be running the test for you. And you see that it is happening all of these for you right now. Look at this. It is already passing all the tests for the valid login credentials, invalid login credentials, empty username and password, malformed username, SQL injections in the uh in the credentials. So it's verifying all of these for me over here. And there are failures. I completely agree. There are failures in the test. But you can easily fix these

12:17 tests for you. So uh you can see that before I get to this particular point if you see the passing test over here it is writing code in Python for me uh and you can see that these are the test case it is generating which is amazing but if you wanted to know if there is any failure and what is the reason for the particular failure you can also uh you can also go and have a look at that particular uh issues for you uh over here. So you see that this is the expected uh value in the line number one and column number one but it is failing and there is this trace log of the reason for this particular failure and what is the cause of the failure and how you can fix this particular uh error as well. So I'm just going to say this is the fix. I'm going to go and copy paste that and I'm going to run this particular thing on the chat. So

13:03 basically there is going to be an AI agent chat which is going to help you to regenerate the test for you and fix the test for you if there happens to be any error. So you see that now it has done all of these for you over here. I think it's still missing some of the details uh on the product creation. Um but let's try to run this test and see what is really going on. Look at that. Now the test has got passed for us over here. So basically it has given me the uh the reason for the failure as well if there happens any. So I'm just going to copy the the reason the fixed reason over here and then I'm going to paste it on the agent chat which is this one and then when I run that it is going to fix that particular issue for me and then it is going to generate the test. It's

13:50 going to run the test and it's going to ensure that the test is passing. So this is how you can see that it's it's going to fix that thing for me and I can now uh fix all my tests one by one. So these are the ways that we can use uh this particular uh AI test agent to perform these operation and you can see that uh this is just like one example that I have given but there are many different examples that you can take and you can see how you can use the power of the test sprite uh in order to test your uh application. And now that we have seen one of the demonstration of how you can run the test and how you can see that the tests can be created like custom test like this one like customuh component test that I have created it also verified by creating a test

14:36 component which is amazing. Now once we have all these test we can now go to the all test option over here and we can download the report as well if you wanted to. I think this is the one. So I'm going to just download the report. So it is going to uh generate the report for me and you will be mind-b blown how the report is going to be because it is not just like any report but the report is actually going to hold a lot more information. Look at that. It is going to be an executed summary of the uh the API test that it has done. So uh it is going to show you that it has 21 test cases it has executed and 12 test case has passed nine has failed and this is the impact uh in total while it is doing all these operation and it's also giving

15:22 me a breakdown of the tests uh along with the test code over here. So this is the uh test code it has executed uh and it has failed because of this particular reason and this is all happening for these these the this reason. Look at that. So this report has already got all the information that is needed for us to do uh the testing of the application which is even more mind-blowing. And the next mind-blowing thing which I wanted to really show you is what if I wanted to check in this entire code that it has generated because you see that the moment I expand it it's going to show me all the test cases and there are failures and passing test cases which is fine but if I wanted to uh check in this code for my future purpose then I can just click this three dot. You see that we have got something called as export

16:08 to GitHub. The moment I click that particular link, it is going to export everything for me in the GitHub and there is this link over here. So basically you need to link uh for the first time and I have logged in uh this particular test sprite with my GitHub account. So it's automatically linking it and you see that now the test sprite bot is going to create those tests for me and it is linking things for me and you see that these are the test cases that I have got which is which is amazing. So all these test cases that I have got over here is been checked in for me this time which is even more better because now we have report we have the test being checked in and now we can run the same test directly on the CI/CD pipeline so that your test is just going to run like any other test that

16:54 runs on the pipeline which is amazing and all this code is written in the Python so we don't really have to worry about it so they generate the test in Python as well as in JavaScript and what I have shown is the Python for now right and the Last feature that I want to really show you is the monitoring feature. So basically uh you can also monitor whether test execution really happens uh all the time without you being within the computer all the time rather you schedule the test execution run the test and it is going to run on the prescribed time that you have defined. So you can uh just go to the uh to the monitor over here uh create a new schedule. You can choose hourly, daily, weekly or monthly. whatever that you want to do it, you can just do that and then you can choose it and then that's going to run the test for you and you'll

17:40 also get a pretty nice report once the test execution is been generated. I'm going to quickly show you how the report looks like. So if I'm just going to open my mail over here, uh look at that. So this is the report which is which has been created by the test sprite and it's going to give me all the information of the test executions over here and you can click this view all button there. And it's going to take you directly to the test sprite and it's going to show you the information even in more detailed fashion over here. See how the execution has happened and things of that nature over here. So this is this is amazing. So you can do create the test cases from the end to end fashion starting from the creation of the test case to the report generations and also running in the pipeline everything the power of AI agent without writing even a

18:26 single line of code. That's the test pride for you. Hope you really like the tool and I'm quite excited to see these kind of tools are emerging with the AI agent powered and we have already released this AI agent courses just recently in Udemy and I can see that the potential is already mind-blowing with the power of these powerful agentic multiAI agent system which is amazing. Once again, thank you so much for watching this video. Catch you in the next one with the UI testing features of the test