Full transcript
0:03[Music]
0:14[Music]
0:35Hello everyone.
0:50Hello guys. I hope everybody is able to
0:53hear us. Still people are joining. Can
0:55you give me a quick? Yes. Hi everyone.
1:02Yes.
1:05So, okay, we'll wait for some time so
1:08that everybody is able to probably hear
1:11us.
1:25Yeah. Okay. So, Udemy, sir, Udemy coupon
1:29not working. Use August 03. Okay. So,
1:33you'll be able to see the Udemy coupon.
1:48So, how was the previous session guys?
1:50Everyone, I hope everybody's attending
1:52the session continuously, right?
2:01Yep.
2:10Okay.
2:21Okay. So I think uh just hit like
2:23everyone so that we start our session.
2:28uh we'll talk about the agenda today
2:30what we are going to cover and all right
2:32so as you all know that
2:35specifically auto automated PR code
2:38reviews with AI this is one of the most
2:41important thing that every developers
2:43will be specifically doing right uh with
2:45the help of n we will try to show you
2:47that how you can do this okay so from
2:51LinkedIn also I think everybody is live
2:54let me just go ahead and check it out
2:57quickly Okay.
3:04Yeah, we're good enough. Like people are
3:06able to see us. Great. Uh I was not able
3:09to attend the previous session. Do we
3:11have the video? Yes, in the live
3:12sessions you'll be able to see all the
3:14videos itself. Okay,
3:18great.
3:22Hi Kish, the rag boot camp Udemy are
3:24very very awesome. looking for end to
3:26end projects. Today only the end to end
3:27projects has been uploaded already. You
3:29can go ahead and see this PR code means
3:32pull request you know so we'll we'll
3:35talk about it as we go ahead you know
3:37you'll be able to see each and
3:38everything.
3:39Okay now before I proceed I really want
3:42to make a quick announcement uh we have
3:45a new upcoming batch which I think none
3:48of you should miss. So this is the batch
3:50details. Let me quickly show you. Uh
3:53this is the batch that we are coming up
3:55with a generative AI for everyone
3:57professionals and leaders. Okay. And uh
4:01the main aim of this specific batch is
4:03to talk about AI how you can use AI for
4:07every professionals out there. Everyone
4:08in short whether you are a student,
4:10freshers, you are a leader working in a
4:13specific company you can definitely go
4:16ahead and check out this specific
4:17course. Now this course is actually
4:19starting from September 14th, 2025 and
4:23uh this is 13th but uh I'll update this
4:25date to 14th and sessions will be on
4:28Saturday and Sunday 8:00 p.m. to 10 p.m.
4:31Along with this uh you know there are
4:33lot of things like who will be the
4:35instructor? Myself and May will be the
4:37instructor. Who is this course for?
4:39sales, marketing, HR leaders, consultant
4:41and entrepreneurs, operation managers,
4:44all leader, decision makers, product
4:46manager, technical professionals,
4:48everyone in short. Here uh what we are
4:50specifically going to master right all
4:52these specific tools you know here we
4:55are not going to focus more on coding
4:57but in understanding how AI can be used
4:59in developing applications that is what
5:01we are going to go ahead and cover up.
5:03If you go ahead and click on enroll now
5:05you'll be able to even see the detailed
5:07syllabus. So let me just go ahead and
5:09click on this.
5:11And once you click on this
5:33uh hello sir, I think you're not
5:35audible. Can you check once?
5:43Hello
5:44sir, I think uh you clicked the button
5:46on the mic mistake
5:48>> just now. Huh? It is
5:50>> or now it's now it's audible. So two 2
5:52minutes last we last we lost you like
5:55the voice only.
5:56>> Okay. Okay. Perfect. So now guys can you
5:59give me a quick confirmation if you're
6:00able to hear me out?
6:02>> Yes.
6:04Let me quickly check. Still not able to
6:06hear.
6:08Just give me a quick heads up. Yeah.
6:11>> Audible now. Audible now. Yeah.
6:14>> Okay. So, perfect. Uh so here you can
6:16actually see guys u we are launching
6:18this amazing batch that is called as
6:20generative AI for everyone professionals
6:22and leaders. Uh this batch is all about
6:26how you can use generative AI in your
6:30projects in your applications many
6:32things as such. You need not even have a
6:35coding background. That is our aim over
6:38here. So for everyone we are trying to
6:40make them understand. There is no
6:42prerequisite as such here. What we are
6:44focusing your on is that once you click
6:46on enroll now, right? So once you click
6:48on enroll now, you will be able to see
6:51this entire page. From this page, you'll
6:54be able to download the detailed
6:55syllabus. Okay? And if you have any
6:58questions, you can contact our
6:59counseling team. From this syllabus
7:02you'll be able to see that you have all
7:04the information out here. It is a 2 to
7:072.5 months program designed for
7:09technical and nontechnical professionals
7:10leader all across domain. Here what we
7:13are focusing on here we are easily
7:15focusing on no code AI tools that
7:18professionals and leaders can
7:20immediately apply to enhance
7:21productivity decision making workflow
7:23automation. You will master platform for
7:25content creation, data analysis,
7:26research and process. Each and
7:28everything we are going to cover with
7:30respect to the tools.
7:32The main aim is that how by using no
7:34code platform we are going to automate
7:36all the specific things right. So we
7:39will learn about generative AI basics,
7:41prompt engineering, no code automation,
7:42AI ethics and governance, data analysis
7:44with AI, AI workflow optimization, AI
7:48security basics and change management.
7:49Right? So all these things is basically
7:52getting covered up. You will be able to
7:55see that we will cover generative AI
7:57fundamentals, AI tools mastery,
7:59professional AI tools, no code
8:01automation, advanced techniques, AI
8:03workflow development and all. Right?
8:06Along with this, we have lot of projects
8:08that we have planned like how we are
8:10taking this specific project. This is
8:11just a basic project itself. Right?
8:13Similar this kind of projects we will be
8:16implementing as we go ahead. Okay. So as
8:19said all the information is already put
8:21up in the pinned comment. I hope
8:23everybody is able to see the pinned
8:25comment or not. Please do let me know.
8:28So if I go ahead and see over here you
8:31should be able to see I have put the
8:34link over here and I will also make it
8:37pinned. Yeah. So it is pinned right. So
8:39if you go ahead and check it out you can
8:41check it out over here and you can even
8:43call our counseling team in case of any
8:45queries. Okay. So this was the quick
8:47announcement. Now it's time like if you
8:50have any kind of problems, any kind of
8:53questions, you can definitely uh ask us
8:56while we are teaching the live session.
8:58Okay. So everything is available. You
8:59can just go ahead and check it out. I've
9:01given the page link over there in the
9:02print comment itself. So now Mike, I
9:05think uh we can go ahead and make them
9:07understand what is the today's session.
9:09>> Yeah,
9:09>> sure sir. Sure. So cool. Everyone I
9:13think we can just start sharing my
9:14screen. Let me know if you face any
9:16issues or you're able to see the same
9:18properly.
9:20So let me just add the screen
9:28wait.
9:31So yeah everyone I hope uh okay now it
9:34should be visible right I hope the same
9:36is visible now. So this was the course I
9:38was talking about. I will share the link
9:39again as well so that all of you can
9:41easily access it right
9:44or you can reach out to our counseling
9:46team as well. The idea is that only and
9:48as I've said the major idea with which
9:49we are coming for this course is that
9:51even if you are someone that hey you
9:53don't know coding or hey I don't know
9:55coding rag or MCP but
10:01I want to use a okay so much like how we
10:05use things around us without having a
10:06in-depth knowledge but of course the
10:08same is still helpful and instead of
10:10going around and taking this particular
10:12path the whole aim is that we are able
10:15to teach you via these different
10:16different tools right that how exactly
10:18can you master these uh overall things
10:21right so that is the whole idea everyone
10:23and yes you can surely reach out
10:25understand AI and what's happening
10:26without coding or technical jarens so we
10:28will be telling you how you can uh use
10:30MCP use rag use AI into different
10:32different everyday tools right we will
10:35try and help you to automate all your
10:36tools okay so you will see that how by
10:39the end of this course you will be
10:40saving a lots of time because we will
10:42make sure that many things which you're
10:44using be telegram slack WhatsApp app,
10:46Google Sheets, Google Drive, then
10:48GitHub, right? All these things we will
10:50help you to easily easily just make sure
10:52that you can automate them. Okay? And
10:54for sure if you easily just follow the
10:56course every week, you will be able to
10:58save 15 hours for sure because we will
11:01be automating your email, your most of
11:03the work, anything which requires just
11:05your little bit oversight and see that
11:07how AI is kind of getting entered into
11:09each and every of these things, right?
11:11So yes, Shahul, we will be using Tableau
11:13as well. So yeah that will be used for a
11:15research based agent. Okay and of course
11:17the basics will also be explained. Okay.
11:19So with that everyone now let's move
11:20forward and today what we are going to
11:22do the overall aim for today is that we
11:26will be now creating a GitHub PR uh
11:30automated automation agent right so
11:33pretty sure that most of you will be
11:34aware about GitHub. So this is the
11:36GitHub platform right in which we can
11:38have different different projects. So I
11:40can show you let's say one of my
11:42projects which uh we have worked on
11:45right. So this is again just some
11:46project which I have created for the
11:48purpose of demo. Now whenever someone
11:51adds a change right so in a company what
11:54happen is that whenever there is a
11:56change which is getting added into the
11:58codebase. So let's say we want to make
12:01it a little bit better in terms of
12:02logic. We want to have some new endpoint
12:04added right then you can just easily
12:08add. So this is just a fast API example.
12:10Now you can add your endpoint. You can
12:12make sure that you are easily able to
12:14use the same. So the idea becomes that
12:16how can I right how can I then automate
12:20any issues or anything is there. So just
12:22I would like to show you a quick
12:23example. Let's say I create a new
12:26repository here with change. Okay.
12:30And in this I just go and I go to main
12:35and I just change it. Right. So let me
12:37edit the same. I'm doing directly
12:39everything on GitHub only so that it is
12:40very easy for you to follow. Now let me
12:43copy this and let me paste it. And let's
12:46say I make a new overall endpoint
12:49available which is V1. And I am also
12:52doing some issues here. I'm making some
12:54changes. Return
12:56uh yes. Right.
12:59And let me just commit these changes.
13:01Okay. So update changes. I can just say
13:04new end point
13:07end point added right. So let me commit
13:10it. Now in a company setting whenever
13:13multiple people are working so of course
13:15there will be different different users
13:17which are working. Right? So on a
13:18particular code base there can be
13:20different different people who are
13:22working and can make the changes. So
13:24let's say it's a team of four where you
13:27are a manager or you are someone who is
13:29in responsibility of handling that how
13:31exactly the changes and everything is
13:33happening. So in that setting what we
13:35want to do everyone
13:38this change which happens. So let me get
13:40the GitHub icon as well. I think it was
13:42available
13:44or some codebase icon. Okay no worries.
13:46So I can just get yeah this is the one
13:48right
13:51whenever someone get this let's say your
13:54codebase so people will have your
13:55codebase and on top of that code base
13:59one can make a change now whenever
14:01someone makes a change
14:04just like how we have updates in our
14:06application and everything we can or
14:09someone will have to review the change
14:10so that it doesn't fail anything right
14:12it doesn't fail your whole codebase and
14:15for that only what we do is I can now go
14:18back to my repo and I can show you that
14:20we can just compare and create a pull
14:22request. Okay, so let me just create
14:24this pull request which is what is known
14:25as PR. Okay. Now this PR as you can see
14:32it has some changes right. I can see
14:34that okay these are the changes. Now
14:37typically what used to happen everyone
14:39is that I as a developer I will come and
14:41I will see that okay what are the
14:44changes are they good or bad. So for
14:45example I will be telling that hey uh
14:48maybe you could have used some other
14:49name for this messages v1 and messages
14:52is a little bit uh confusing right? So
14:55we already have this messages. Then why
14:57is this yes here? I think this can be
14:59avoided. Right? Then we could have added
15:02some documentation here as well. Right?
15:04So those are the different different
15:05kind of changes which we can have here.
15:08Okay? So with respect to that what
15:11happened either you will be spending
15:13your time in making sure of this two and
15:16fro between. So let's say this is the
15:17user one who got your codebase right. So
15:20I will have this user. He got your
15:23codebase and he made the change. Now you
15:25being a manager, let's say you being a
15:28manager, let me just copy this
15:31from here. And you are the manager here
15:34or let's say tech lead. You will be
15:37telling that hey please do this change.
15:40Please make sure that you handle this so
15:42that this doesn't break. In that way
15:44what we are trying to do we are trying
15:46to now make the manager work easy or
15:48overall the team work easy altogether
15:50that hey all these changes they can just
15:52get a prevet by AI where it will be able
15:55to tell that what exactly are the things
15:58which you need to handle. Okay. So AI
16:01only will be telling the basic level of
16:02changes, it will be checking the logic.
16:04Most of the I would say 80% of the
16:06things AI can directly help you to do
16:08here and you will be pretty good to
16:10easily just uh then make sure that you
16:12do those changes and easy just then move
16:15forward and make make sure that you are
16:18focusing on the major changes there. Now
16:21the same is very very helpful like we
16:23use the same into our startup as well
16:24where it has been integrated and for the
16:27first year second year or at least the
16:28intern it helps me to save a lot of time
16:31earlier I have to go and see each and
16:33every file I have to add my commits for
16:36the basics of the changes as well I have
16:38to do those things now the same all has
16:40been automated right and a very very
16:42good product on the same thing uh I'm
16:44not sure how many of you must have heard
16:45of it so there's a code rabbit right
16:48which is a pretty good similar kind of a
16:50AI code reviews. So they have created a
16:52whole uh thing here, right? So if you
16:54will see what code rabbit does is
16:56exactly the same thing, right? It kind
16:58of a PR is created. It adds that hey
17:01what all is the difference here. So they
17:03are selling it right. It's a pretty good
17:05thing. Uh they are selling the same as
17:06you can see for prices and it does the
17:09similar kind of a thing. Of course with
17:10a lots of uh training lots of better
17:13approach it is able to do those things
17:15and it will be adding the similar kind
17:17of comments. So this same thing we are
17:18trying to now make sure that we can do
17:20into our workflow. Okay. So that's the
17:23whole plan and this is the problem which
17:25we are solving everyone that whatever
17:28changes happen and I can maybe just have
17:30a another icon right so that you can
17:34understand it pretty nicely. So let's
17:36say whatever code I am having here,
17:38whatever code I'm having. Now whenever
17:41this code changes, so if this is
17:44v_sub_1, then I can just have another of
17:48this file which will be v_sub_2.
17:51I say that this is my PR getting
17:55generated. So in the technical sense, it
17:57creates a PR, right? in the technical
18:00sense it will be creating a PR
18:03and I have to make sure that I can just
18:07handle all the changes everything into
18:09this PR. So let's see how we will be
18:11automating the same with the power of
18:13AI. Now doing this without any uh no
18:17tools or with uh code is also possible
18:20but that is a very very difficult
18:21process right and this is how we are
18:23going to uh the problem is now clear now
18:26we will see that how easily we can just
18:28integrate and use the same okay so yeah
18:31that was the problem everyone anyone has
18:33any issues anything which is not clear
18:34let me know in the chat and one more
18:37thing so I was just trying out one very
18:38very new thing how many of you are aware
18:40about uh Google uh this thing Google
18:43Nano Banana
18:46I'm working on the same as you can see.
18:48So it has created few of my images from
18:50a single image. This is me in 1990s.
18:54This is me in 2000s I guess. This is me
18:56in 1970s. So this is pretty funny
18:59though, right? And uh this is me with uh
19:03Okay, let me show you. So let me know if
19:06you would like a video here as well. And
19:09as you can see this is me.
19:13Okay, let it load. So meanwhile you can
19:15tell me if any issues are there in the
19:17understanding or and I will show you
19:19that. Okay, see this is me with Elon
19:21Musk and Jeff Bezos. So this is me with
19:24Jeff Bezos on Amazon and this is me with
19:28me with Jeff Bezos and Elon Musk. Okay,
19:32so I hope you like this as well. We will
19:34surely be going and seeing this. So
19:35we'll be uploading the video for the
19:37same in on my channel for upcoming days
19:39and how you can use this into your
19:41application. So this can be a very very
19:42big business opportunity right. I could
19:45just easily now say uh
19:49can you make the middle guy
19:53wear a leather jacket and let's see okay
20:00so along with how you can use it here I
20:02will be showing that how can you use the
20:03in the code but in meanwhile do tell me
20:05if this use case or the problem
20:07statement is clear everyone then we can
20:08move forward with this right
20:13so hopefully it will help me wear a
20:15jacket
20:16so that I also look a little cool with
20:19Jeff Bezos and Elon Musk and yeah anyone
20:22has any problem now in the in this
20:24particular
20:26use case which we have discussed
20:28so let it generate meanwhile
20:37see
20:38right
20:40cool so video will be dropping on my
20:42channel soon I think by tomorrow let's
20:44get back and see the same so the idea as
20:48I said will be that we are able to use
20:50it directly we are able to create a
20:51workflow in which we are going to
20:53automate this okay
20:56tell very quickly everyone if there is
20:57any issues anything is just not clear
20:59and I shall help you with that
21:04any other doubt as well we can discuss
21:06because please stick with the topic so
21:08def coder I'm asking the doubts
21:10meanwhile I'm showing this I'm on the
21:11topic only you see this this is the
21:13topic Right? You should know about the
21:16AI updates.
21:19Do we have a job out there for no code
21:21tools that we will learn in this course?
21:22Uh Raj, uh many companies are now kind
21:25of reaching out. They are looking for
21:26people who can help you automate. We'll
21:28just give you an example. So I was
21:30recently reached out with a company and
21:32they wanted that these tools can help
21:34them just switch from their previous
21:36automation tools which are inhouse
21:38because they are having a good support
21:40of AI. So again if you think that okay
21:42just by learning a tool you will be able
21:44to get that might even not the case but
21:46yes if you are in let's say non techch
21:47and you learn these tools your use case
21:48will be better and you will be able to
21:49do things better and yes there's an
21:51opening for people who understand these
21:53things right same for tech as well so
21:56companies are looking forward and
21:57majorly in starters etc if you can
21:59create and explain these workflow easily
22:02create with a common use case then sure
22:03the same will be very helpful okay
22:08clear everyone
22:17Great. Uh I cannot automate the capture
22:20to bypass capture for scraping data. Is
22:22AI help? Uh defcoder I think you will
22:24have to explain this a little better.
22:25Like if you want to just uh get through
22:27the capture if it's a single uh let's
22:30say image related capture you can use
22:32AI. Otherwise it might be difficult
22:34because now those captures are also
22:36getting a lot smarter because again
22:38that's the case with the right
22:42for any coupon related thing anything
22:44which is not clear please reach out to
22:46the number so that shall be helpful okay
22:52great so since uh everything is clear
22:55and we have understood the problem
22:56statement in depth let's now see how we
22:58will be doing it. So for the same again
23:01we are going to use a no code tool. Now
23:04again there are many many no code tools
23:06in the market like NA10 make relay
23:08Microsoft power automate Zapier the one
23:11we are going to use is NA10 and I think
23:14that all of you have been following the
23:16previous videos as well and know why
23:18this is very powerful and very useful
23:20right so for this let's go to any now
23:23you could easily just go and click on
23:24this get started I will share the link
23:26with all of you so you create the
23:29account the workflow which we will be
23:32using here right everyone is already
23:35been uploaded so I think uh the team
23:38will share that from where you can
23:39download the workflow so to make sure
23:41that each and every one of you are able
23:43to follow the same easily we have made
23:45sure that we already upload the uh
23:47workflow pre-created so this is going to
23:50be our workflow let me show that uh one
23:53minute
23:56so I can just create new workflow
24:02and I will show you that what exactly we
24:04are going to do there. Okay.
24:06So 1 minute.
24:11Yeah. Uh okay. Yeah. I think this is
24:15going to be the workflow everyone which
24:16we are going to use today
24:19right and I will walk you through each
24:21and every of these steps. So since you
24:22will have the workflow already I think
24:24the same will be very very easy to
24:26follow.
24:36anything meanwhile
24:39I will just share that from where you
24:40can download this particular workflow.
24:42So if you will go and
24:45let's say just enroll onto the free
24:48course there we have uploaded so as
24:51there is a single place where all these
24:53different different workflows are
24:54present. Okay so I will just show you
24:55quickly where exactly the same is.
25:01Okay, one minute. I think I copied a
25:04different link.
25:16Okay.
25:29One minute everyone. So just let me get
25:31this link so I can share with all of
25:32you. Yeah. So this is the link everyone
25:35which I'm sharing. If you will enroll
25:37for free on this course, you will see
25:38the previous videos and everything as
25:40well. Okay. So sharing this with all of
25:43you.
25:49Great. So now let's move forward. Now
25:51this particular workflow actually it is
25:54a part of an already like it is actually
25:58in the templates of any right so if I
26:01will go on GitHub and let's say if I
26:04search here you will find the template
26:06but I will tell you what exactly is the
26:08problem there. So if we go on GitHub and
26:10let's say we choose PR review.
26:16So this is again an awesome template
26:18like the same template which has been
26:19given but this particular template is
26:21having a problem. I will show you that
26:23quickly so that you don't fall for this
26:25uh issue. I think they it got uh a
26:28problem because of the new uh updates.
26:32Now when you use this pre-created
26:34template right when you use this
26:36pre-created template and let's say when
26:39you just try to execute this I will
26:41quickly add my credentials etc. So pay
26:44attention here everyone what we have to
26:46do we will be now authenticating right.
26:49So this NAN NAN we will be connecting
26:52with our GitHub just like how we do it
26:54for other different different
26:55applications we have a GitHub account
26:57right and in this GitHub account we will
27:00be now using NA10. So there needs to be
27:05this connection via which we can easily
27:09make sure that N8 has access to each and
27:13everything on my GitHub. Now for the
27:16same it provides a direct connection
27:18where we will be using OOTH2
27:21authorization right. So let me just
27:23write it like this and we will be
27:25creating a connection like this. For
27:27this it is asking for a OOTH2
27:29connection. We could have done the same
27:31via access token as well but normally I
27:34think we all know that how OOTH2 is a
27:36little bit easier to do right ooth 2 is
27:38a way way way easier to do so. So for
27:41that let's just go to ooth 2. Now if you
27:44will just click on connect my account
27:46make sure that you are already logged
27:48into GitHub right? So if I will go onto
27:50this browser on github.com
27:54then you will see that how my account
27:55the one which I want to connect is
27:57already open. So this is my account
27:59everyone as you can see my photo here
28:01right. So once we have the GitHub uh
28:04logged in just go to the workflow again
28:07and just click on connect my account.
28:09Once you click on connect my account it
28:12will try to open and it is showing that
28:13connection is successful. It will just
28:15ask you few things right. So what I can
28:18do let me go back. Let me go back. Let
28:22me show you by creating another
28:23credentials. This time let me click
28:25connect my account
28:27and okay so let me delete I think I've
28:30already connected this thing so I can go
28:33on
28:34apps so
28:37your repository settings
28:42applications so I can just easily see
28:44that how can we reduce this so
28:46integrations I think lovable is
28:48integrated
28:49and o side I have this any so let me
28:52remove this let me revoke
28:55so that we can reconnect and I can show
28:56you the same from scratch. Okay.
28:59Again uh this will be now invalid. So
29:02let me just create a credential again.
29:04This time when I click o to connect my
29:06account it will be the way you will see
29:08right. So let me click this. And now see
29:11you will see this kind of a window. N by
29:13N10 wants to access your may 953
29:16account. So my GitHub is
29:18github.com/mayank953.
29:20Now when I click on authorize and
29:23integration the same will just happen.
29:25Okay. So it's that simple
29:27and I can close this. Now the same has
29:30been authorized. Now once we have
29:32authorized this everyone I can just
29:34select the repository owner owner. So
29:36here you have to specify the name
29:38exactly right. So if I will write may
29:40953 then it will be able to pick that
29:43up. So it is just a open search box. It
29:45is not going to get it for you. And I
29:48can just click on this button now to see
29:51that yes my account is getting opened.
29:53Okay. So in that way you will be adding
29:55your username. So GitHub github.com/
29:58whatever your username is you will be
30:00adding that username so that you can
30:01easily connect. Okay. Now once that is
30:05done then next comes the repository. Now
30:08I am showing you all of these things
30:10with this NAT demo PR. NAN demo PR. So
30:14let's do one thing. Let me just go here
30:16again. Let me close this. Go here and
30:19search for NAN demo PR.
30:25And now when I try to execute this, you
30:29will see that some error is coming. So
30:30let me first resolve the outstanding
30:32issue. So let me disconnect this. I
30:34think this has been disconnected, right?
30:37Let me open this again. Now when I try
30:39to execute this, see it is showing you
30:41an error. Cannot read properties of
30:43undefined reading HTTP code. Okay. So
30:47what I wanted to show you was this only
30:50that this particular
30:52uh workflow or template which is present
30:55that is the template which we are also
30:56trying to use but the same doesn't work
31:00and the reason that it doesn't work you
31:02can see right on your screen that how it
31:04gives you this error. Okay, it gives you
31:06this particular error. So for the same I
31:09want that you download the workflow by
31:12uh which I have provided to you so that
31:14you don't face this error when you
31:15directly try to launch from the overall
31:18uh let's say this uh this template.
31:21Okay. Now if I show you the one which I
31:23have given. So let me connect my account
31:25again.
31:27I think it will be able to connect
31:29easily. Yeah. So now let me go back and
31:32this time I'll run this. Let's see
31:35problem. Okay. Let me resolve the
31:36outstanding issues. I think I will have
31:38to disconnect this. So let now open it
31:41and let me see.
31:44So okay I think here as well we are
31:45having this issue. Let me quickly check
31:47everyone what exactly we have to fix
31:49here.
31:50Right.
31:52So
31:54I will just make sure that I can
31:58get the one which you all have shared
31:59because that is totally working without
32:01any problem.
32:03So this is the one I guess right and
32:07what I can do here I can just do one
32:10quick thing. Delete this one and go
32:13here. Sorry for GitHub and add a trigger
32:17on a pull request. So this trigger is
32:20the one which we have to add and I can
32:23load it here or delete for the timing.
32:26Now let me do one thing. Let me this is
32:28connected. Let me again add may 953.
32:32So this is the one and n demo
32:37PR and now when we execute
32:41okay this I think has been saved now let
32:43me execute it again
32:46cannot read properties okay let us fix
32:48this thing meanwhile uh we see this
32:50everyone I hope all of you are getting
32:52this error let's see that if you have
32:54any doubts let me uh tell me please if
32:55you have any doubts whatsoever
33:02So workflow file uh nimad will be there.
33:04It will be I think uploaded with a g uh
33:07with this link right with a overall
33:11uh drive link. So drive.google link. It
33:13will be uploaded there. You can just
33:15check it quickly.
33:33Tell me are you able to uh come to this
33:36point and face this issue? What exactly
33:37is happening? Because I think everything
33:39is given to you already.
33:43How to go with level as codebases
33:45shared? So Jag again uh if you are
33:48having the authorize code base you will
33:50just need to connect that to any okay
33:51but your enterprise may have okay it may
33:54have some way some let's say things
33:58where you cannot connect to it right you
34:01uh so and uh J what happen is that it
34:04has an access to your private
34:06repositories as well so nothing will be
34:08there on the open okay so that should
34:09not be the case in that way you can it
34:12is all safe because you can just have a
34:14private repository So it should not be
34:16out. Okay, that should not be the
34:18problem. So that way it is safe. That
34:20should not be a problem.
34:34Tell me everyone are you able to follow
34:36easily? Are you getting this error?
34:41The workflow is there. I have just
34:42checked uh in the setup as well. Let me
34:45quickly just add it.
34:48I see the workflow resource where it is
34:52there. Let me just add it again so that
34:55you can uh use it. Right.
35:10Cool. So everyone just check it again. I
35:12think I have added the same. Shall be
35:14fine now. Just check and let me know.
35:17And also let me know if you're facing
35:18any issue. I want that. Okay. V this
35:20live you all are also doing the same
35:21with me. Right. I don't want to just do
35:23it uh like myself. I already have done
35:27this. But N should be a third party
35:29tool. Uh well J for that what you can do
35:32is you can self-host. So Nin can be
35:34self-hosted inside your uh thing. So any
35:38self host
35:40right if you host it inside your
35:43company's uh let's say VPS or anything
35:47then I think it will be all good right
35:49so easily that can be done
35:55cool everyone I hope that all is good
36:24Tell me in the chat quickly
36:27as you said the live we are doing so
36:29that all of you can follow right
36:31>> mind do you have that workflow link the
36:33materials people are asking for it
36:36>> uh what sir
36:38>> the materials uh people are asking for
36:40it I'll add
36:40>> yeah so sir that has uh that has been
36:43already added in the dashboard via drive
36:44link so I think each and every one of
36:46them should be able to access it.
36:48>> No, no, no. Actually, uh you just send
36:50me that link once. I will upload it.
36:52>> One minute. 1 minute
36:57shared. So, I have shared the link.
37:06Okay. I think the same. So, yeah.
37:10>> Cool everyone. So, we have done this. Uh
37:11let me know if you are facing any issues
37:14so that we can move forward with this.
37:16And yes, for people who are asking, we
37:17can easily host self-hosted nit like
37:20onto our own architecture infrastructure
37:22sorry. So that should not be a problem.
37:23Okay,
37:28let me know if you're able to get the
37:29workflow. I want everyone to follow
37:31alongside with me and in few seconds you
37:33will see that how nicely we're able to
37:34do all these things.
37:51Meanwhile, if anyone has any question,
37:53they can ask the same regarding course,
37:54regarding AI in general, regarding
37:56anything. If you want to have my
37:58orchestr views or anything, we will be
38:00happy to clear that
38:21workflow. So I think yeah people are
38:22able to get the workflow.
38:28importance of genai nowadays. Uh I think
38:31AFK a lot. The reason is that no matter
38:34in which field are you in jai is a kind
38:37of a skill which is sought after plus if
38:39you know how to use it. Now I know that
38:40many people might be thinking of these
38:42no code tools and everything maybe as
38:44like okay they're not useful and
38:45everything but believe me I personally
38:47have seen that if you use the same you
38:48are able to save a lots of time plus you
38:50can generally automate a lots of things
38:52right. So that is the case and yes in
38:55that note it is very very useful.
39:03Great everyone. So
39:05I think we all can move forward. Now
39:08once you click on this you see that how
39:10we are getting this error. Right now
39:12this is actually one of the error which
39:14have been introduced again in a similar
39:17manner that someone must have put up a
39:19change. Right? Right. So I'm just
39:21searching it on Google and someone must
39:22have put up a change into the enit
39:24codebase just like how we are discussing
39:26about a PR related change. Right? And
39:29actually this error got induced. Okay.
39:30So you can see kit trigger cannot read
39:32properties of undefined reading HTTP
39:34code. So if you are a forum kind of a
39:36guy who can read through this the way I
39:38read through this and was able to solve
39:39this error. Uh you will be good. But if
39:42not then of course I am the person who
39:44can help you with this. Right. Right. So
39:46there again they are saying that hey uh
39:48some issue is there so how to solve it
39:50everything. Now they have created a pull
39:53request for this. If you will see they
39:55have created a GitHub issue for this
39:57where also they are having a pull
39:59request. So you can see the pull request
40:01here and the same thing which we are
40:03kind of trying to do error HTTP code
40:05they did this changes and I think in
40:08that something changed. Okay. So that is
40:10the whole problem and yes that is the
40:13reason which by which we are also
40:15focusing on how can we use the same uh
40:17properly. So now what we will do
40:19everyone pay attention. So let me close
40:21this. Let's go back right. Let me just
40:25leave this. Let's go back and I'm going
40:29to save this workflow. So you will be
40:31easily able to get this particular
40:32workflow via the JSON which we have
40:34uploaded. Now I'm going to create a new
40:37workflow. So pay attention. I'm going to
40:38create a new workflow.
40:41Now once you create this new workflow
40:43here do GitHub and in this search for
40:48pull request. So on pull request. Now we
40:52will try to do the authentication again.
40:54So just do the authentication again on
40:56this new sheet. Just make sure that you
40:58properly do this. So this has been
40:59authenticated now. Now what I will do
41:02pay attention. So now I will again
41:04select myself. So mayank 953,
41:08right?
41:10May 953 N8N.
41:15So what was this exactly? N8N demo PR.
41:19And now when I click this, let's try.
41:22See this time it is listening right? It
41:24is not giving me that error. It is
41:25saying that hey I am listening for test
41:27event. The second you create a PR now
41:31that error has been kind of handled
41:33where it will be now checking that okay
41:35so I was able to understand the same
41:37while going through this particular PR
41:39and everything but for our use case I
41:41think we are pretty good so what I can
41:43do I can just go back and I will just
41:45quickly create a new PR so just uh pay
41:48attention okay I would just like to
41:50create a new pull request let's say I
41:52create a new pull request
41:55that pull request will be base will be
41:57again and I can maybe let's say select
42:00the PR example. Okay, let me now create
42:04a P uh PR. I think okay, this has
42:06already been created. Let me try to
42:09close this pull request.
42:11I just want to show you by creating one
42:13that how exactly our NAT will be able to
42:15catch that PR. Okay, so I will select PR
42:17example. I will create a pull request.
42:21Now once I create this pull request it
42:23will send a signal to my N10 that hey
42:26something happened and what was that
42:28something that something was this thing
42:30that it is able to now easily get the
42:32details. Okay so many many details it is
42:34able to get and this is what we want to
42:36use. Okay. So that is exactly how why we
42:40need to make sure that we are able to
42:42solve it. And the way to solve it,
42:44right? The way to use GitHub trigger in
42:46nitin uh till they solve it is in this
42:50manner only. Now what I can do I can
42:52just go back right I can just open the
42:55full workflow again the one which I've
42:57given you and I will be now just copying
43:00the other nodes. So let me create
43:02another new one. Let me create new one.
43:05new workflow.
43:07Now this workflow gets created. I will
43:10just open. So the file which you are
43:12getting you can import from file or you
43:14can just copy paste that file here like
43:16this. Okay. Now once I have copy paste
43:19uh copy and got the file what I will do
43:20I will just delete uh this particular
43:22thing. So I will delete this faulty node
43:25in the workflow and I will just copy all
43:28of them. So let me copy them go back and
43:31the best part is that you can generally
43:33just copy paste it like this. So see the
43:35whole thing is now together right I'm
43:37able to get each and everything I will
43:39just move them and connect it with my
43:44this particular flow thus making the
43:46overall
43:50now we to fix.
43:59So when the file difference which need
44:02to be worked upon then create target
44:04prompts for PR differences. So here we
44:06are just setting the prompt. Okay that
44:08what exactly do we want it to have. So
44:10for example here again some code is
44:12there you can just move forward and we
44:15can see you are a senior Python
44:17developer. Please review the following
44:18code changes in the files. We will get
44:20the differences. Once we get the
44:22differences then we can just discuss on
44:24the missions. Review the proposed code
44:25changes by file and by significant
44:27modification. Generate inline commands.
44:30Ignore file without patches. Do not
44:32repeat the code snippet or the file
44:33name. Write the command directly without
44:35introducing the context. Okay. So with
44:38that now once we have done this then we
44:40will move forward with this agent. So
44:41this is an AI agent everyone which we
44:44will be using and it will be having a
44:46brain of its own. So in this we can
44:49easily just let's say use and claim this
44:51free credits. So when you create a new
44:53entit account, you can just claim these
44:56credits or if you want you could just
44:58maybe use any other model totally on
45:01you. You can use a Gemini model, you can
45:03use anything totally on you. Okay. So I
45:05have just claimed this free credits and
45:07I am selecting GPT 40 from this. Uh so
45:10let me get GPT4 mini now. Okay. So these
45:13are the models which are available.
45:16Then we will ask or since we are
45:18connected with GitHub we will try to add
45:21back whatever changes the AI is
45:23suggesting the agent is suggesting back
45:25to our overall PR. Okay. So that is
45:27going to be the flow. Now so everyone
45:30tell me now are you able to follow this
45:31properly? I think this is a valid use
45:34case. Each and every one of you shall be
45:35able to do the same. Right.
45:38Let me know if you have any doubts if
45:40anything is not clear.
45:48Grock is also supported foder. So you
45:50just let's say I delete this particular
45:53uh connection. Go here select gro and we
45:58can get the gro chat model. So see that
46:00easy again from the credential go and
46:02create a new credential. Add your key
46:05and yes I think we all are good here.
46:08I will just check if I have the key
46:10here. So let me just check with my key.
46:14save
46:15and seems like the same is working. So
46:17now since this is working everyone I
46:20could just select any of the models
46:21totally for free as well right so the
46:23gro open router all of these are
46:25supported so you don't need to have that
46:26but I think the best part is that how
46:28nit also gives you these free credits
46:30right
46:34so I will stay with openai
46:39great
46:46any doubt anyone Let me know quickly in
46:47the chat if anything is not clear. I
46:49will help you to quickly clear the same.
46:51Again, it's a very very helpful use
46:52case. You can surely just implement the
46:55same in your company and believe me,
46:56your company's going to be very happy.
46:58If I would have had this uh lots of
47:00time, two years back must have been
47:02saved in my life. Okay,
47:08let me close this and just zoom in.
47:17Great. So now since the whole workflow
47:19is clear, okay, what we will be doing
47:21everyone, we will now just try to
47:24understand and just execute this
47:27workflows one by one. So this is a API
47:30which GitHub provides where we can get
47:33via the PR that okay what exactly is the
47:36difference between the files. So let me
47:39uh just end click this and in the table
47:42or let's say via schema only we can
47:44understand that okay this is an awesome
47:46repository regarding any tutorial for
47:48GitHub. Now if I go back to my GitHub
47:51and if I go back to my file changes you
47:54can see that this is something which I
47:55have added in my readme right so readme
47:58+ one. Now if I again go back to my
48:01workflow right this has got all the
48:04files. Now next is create target prompt.
48:08So this code will just try to
48:10understand. Uh for the timing if you
48:12don't understand this code should not be
48:13a problem. Let me just execute this. We
48:16have just given it some prompt and just
48:17asking that hey this is the user
48:19message. You are a senior Python
48:21developer. Please review the following
48:22code changes in the files. So I can show
48:24you the same in JSON. This is the file
48:27which is readme.md. The difference is
48:29this create this is an awesome
48:32repository regarding this and your
48:33mission is already given. Okay,
48:37now I will go to the next one which is
48:39the AI agent. I will just execute this
48:42right. Let's me execute this and this
48:45agent will take up this user message and
48:47tell me that okay what exactly are the
48:48changes which is suggest. So it say it
48:50is saying this message is a good start
48:52but it could be improved by adding more
48:54details. Consider mentioning what N10 is
48:56how this repository can benefit users
48:58including installation instruction or a
49:00brief overtime. Now if you want you can
49:03just give it some more information here
49:05that hey this repository has to do with
49:07fast AI not with MCP. Uh please make
49:10sure that you are properly giving the
49:12same inpoints. So let me just give it
49:13you are a helpful assistant. Please give
49:16your
49:18analysis in points
49:22the repository
49:24has to do with
49:27fast API. Now when I execute this, see
49:31how the output will change because we
49:32have told our agent that hey something
49:34is different. See the read me file is
49:36newly created and provide bc description
49:38description mentioned this which may not
49:40accurately reflect the content if the
49:41repository is primarily about fast API.
49:44Consider updating the description to
49:46better represent the project. Okay. Now
49:49since we have done this let me move
49:50forward and now it is not able to get
49:53this. So let me fix this. We can just go
49:55back and change this to PR review.
49:59The reason was that in the original
50:01workflow so this was PR trigger not PR
50:04review. So let me fix this. Okay let's
50:07go back and let me rename it. So I can
50:10just press space and rename the same PR
50:12trigger.
50:13Now see this has become green because it
50:15is able to get this particular thing.
50:17Now when execute this let's see it is
50:20able to get all the information. It is
50:21able to get the body as well. If I go
50:23into my PR let me go in here. I think
50:26nothing has been done till now. Now when
50:29we go back onto our workflow then the
50:32last one is the add label to PR. In here
50:36when we click this
50:38as we can see this has done the same.
50:41And when I go back to my PR and let's
50:43say refresh this
50:47uh 1 minute I think it was the different
50:49PR. Let me check which PR it was. So
50:52create readme
50:55no conflict commits
50:58files changed this is the file which has
51:00been changed
51:02let's check have we missed out this
51:04thing where exactly it is supposed to do
51:06so let's go back my 953 in our demo PR
51:13so where exactly it has missed it out
51:14everyone let's see the same or let me do
51:17one thing let me just create a let me
51:20just now since we have understood the
51:22workflow. Let me execute this workflow.
51:24So it is not waiting for the trigger
51:25event. Let me go back. Let me go back to
51:29my main
51:31create a new PR change here. So I can I
51:34will just create a new PR with changes.
51:38Okay.
51:40In this PR I will just quickly go and
51:43let's say I will just add a readme here
51:44only. Okay. This is an awesome repo for
51:50N8 changes.
51:52Now when I commit this changes, I can
51:54commit this changes
51:57and I can go back and create a new PR.
52:00So if I just show you, let's say if I
52:03just select 1 minute. So what I can do,
52:07I can just select this try vertically.
52:12So let me execute this. So it listens to
52:15this. See it is waiting for the trigger
52:17event. Let me compare and pull request
52:19right. I will just create this PR and
52:22pay attention. Uh I can actually make it
52:24active from here. So pay attention. I
52:25can just make it active from here. Once
52:27I make it active, right? Then it will be
52:31listening. So it will be just listening.
52:33And if I create a pull request,
52:36so the pull request has been created.
52:38Now when we go to execution, we will see
52:41that how an execution has started. So
52:43this execution is running everyone and
52:45see how quickly it has just added. This
52:47line introduces a new readme which is
52:49crucial. Additionally line introduce a
52:52description of the repository. However,
52:53it lacks the detail and this is exactly
52:56what was given out right. So if you see
52:57here this line introduces a new readme.
53:00It is easily able to now just add it
53:02that quickly. So I hope you are all able
53:04to understand. I think it got uh k up
53:06two times but that should be fine. We
53:08can check why it happens. But yes, as
53:10soon as I create a PR, it is showing me
53:12that okay, what exactly happened, right?
53:14So that is the full benefit.
53:17Now to make it a little bit more better,
53:21what we can do, pay attention again. I
53:23can go to my repository.
53:26Let me know if anyone has any doubts
53:27now. Let me meanwhile take the doubts.
53:33Can you click on the GitHub trigger node
53:35again? I can surely do that. If I just
53:37go outside, I can go on editor and in
53:39here this is the GitHub trigger node.
53:41Okay, everyone see I've just done the
53:43authorization. Uh just make sure that
53:45you start this up in the new one. Okay,
53:47not take that overall uh
53:51example just that will company allow to
53:54use uh n if you learn about how to use n
53:57on a self-hosted infrastructure which is
53:59very very easy via the docker image. I
54:01think right now companies can easily
54:03allow that because again nothing is
54:04going outside from that right. So if
54:06you're and you can generally use AI
54:08which is local hosted as well. So let me
54:10know if you want a video on that and I
54:11will help you with this. Okay.
54:14Can someone pursue career in ji or agent
54:17by only learning this no code way? What
54:18career option might be open for an
54:20individual if the course is taken or
54:21knowing anything? So rupajjit I will be
54:24telling this from an experience of mine
54:26like it started from a technical
54:27perspective.
54:29If you only understand this uh no code
54:32thing and let's say you want to go
54:33through the technical field then it
54:35might be very difficult that said if you
54:37want to now using these tools if you
54:39want to now go into the nontechnical or
54:42semi-technical field that might be still
54:43doable so proper tech I think it will be
54:46a problem because somewhere or the other
54:48you might have to go beyond by coding
54:51right so let's say if you're in sales HR
54:53marketing if you are even let's say
54:55leading a team after 10 plus years where
54:57you don't code but you need to
54:59understand that okay what exactly are
55:01these different different concepts then
55:03the same can be helpful not the other
55:04way around okay
55:07especially in companies like JP Morgan
55:09etc. So again techos I think I know from
55:11where you are coming from Goldman Sachs
55:13where I have worked that was also very
55:14very particular about what exactly are
55:17they using but I believe that if you can
55:20host and show them via their infra and
55:23show that why the same is useful how the
55:25data is not leaving onto a locally
55:28hosted AI then I think they should not
55:31be a problem but your presentation your
55:33way should be like your overall
55:34confidence should be very high like for
55:36example if I have to convince my
55:38managers Now I think I could easily do
55:40that right I could easily easily do that
55:42thing because I know that how it can be
55:46hosted by using a simple uh own AI so
55:50with that knowledge I think I can now do
55:52that but of course uh if you don't have
55:54that then it will be a problem so
55:56showing that how data is not leaving
55:58everything is locally hosted how it is
56:00very helpful it can help to do so many
56:02things then it will be better because
56:04doing the same via code that can be a
56:06problem can I create AI tutor Which kind
56:09of audi zahu pink zahu
56:27great so meanwhile this uh comes now let
56:29me just show you the best example of
56:31this let's say this particular thing is
56:33still happening what we nor do is like I
56:36will be working on this code base so let
56:38me go to main and I will be doing the
56:40same via some VS code on IntelliJ but I
56:42can create let's say may changes as my
56:46new branch.
56:49Now in this branch I will go to fast API
56:51template and I can go to let's say
56:53main.py.
56:55Now I can edit this right. I can just do
56:58the same thing which I was saying to you
57:00in the initial that hey now I can add a
57:02message one v1 something like this.
57:07I can do like let's say if I'm sure I
57:09don't know of coding I do a lot of w
57:11coding it can just print something here
57:14print
57:16hello world right then I can just commit
57:20this changes right
57:23I can go back to my this branch only I'm
57:26making all these changes right so let me
57:28go to the branch let me just go to
57:30models py let me make some edits here
57:35and I can again do some class
57:38not usable or error here right error
57:44message string I can just add this as
57:47well right now I have done few changes
57:50and in this particular uh PR I can just
57:53now create a new PR okay so let me just
57:55create a new PR so my changes let me
57:58create a pull request and let's see the
58:00same in action now see I've created a
58:02pull request
58:05each and everything is happening So let
58:06it run. Checking for the ability to
58:08merge everything. Right?
58:12I have two commits, two changes. But
58:14see, as soon as I created my PR, right,
58:17I have not done anything. I have not
58:19changed anything. My workflow is
58:21running. The workflow is pretty active
58:22in the back end. It is able to tell me
58:26that hey, the new route introduces a new
58:29API endpoint. This is significant
58:31addition to API but naming of the
58:32endpoint could be improved for better
58:34restful uh structure and clarity. The
58:36function add message now contains logic
58:38to generate unique ID. The approach of
58:40using this works but could lead to
58:41issues. So the addition of hello world
58:45right you can see here the addition of
58:46hello world that also it is able to tell
58:49that how it is not very helpful. Right?
58:51So let's see the same the addition of
58:54hello world is unnecessary for
58:56production code should be removed. It
58:58can clutter logs. Consider using
59:00structured logging if needed. The
59:02message payload model now includes a
59:04message ID. So I hope you can see that
59:06how it is telling each and everything
59:08right in front. Right? The error class
59:10is not defined as a pyantic model which
59:12should be the case. So now think that
59:15even if a fresher or someone who doesn't
59:16know the code, right? How easy it is
59:19like normally it is giving you these
59:20messages right and this is the benefit.
59:23I could have the first level of checks
59:26very very easily done. And this is again
59:29what this code rabbit kind of does in a
59:30hindsight that you can easily create
59:32this then it tells you the issues
59:34everything going a lot inside your code
59:36right but with respect to this
59:38particular thing it can easily get so
59:40many things for you considering adding a
59:42response model URL can be benefit this
59:45time should be removed that is does not
59:46fit within the fast API application so
59:49all these things is directly being
59:52happening saving me a lots of time now
59:53of course you can have a more complex
59:55use case more complex changes is and
59:58everything will be handled. Okay. So
1:00:00yeah with that we have uh just completed
1:00:02this. If I go on the execution you can
1:00:05see that how the execution happened
1:00:06right.
1:00:09How the execution happened. It is being
1:00:10completed. If I make it inactive now it
1:00:13will be fine. So yeah I think this is
1:00:15where the overall power of our GitHub AI
1:00:19automated PR comes into picture. We have
1:00:21also seen that how can we correct the
1:00:23error which we normally get. How the
1:00:25same is useful in real life. As soon as
1:00:26I create a PR the same gets triggered
1:00:28and it can just handle it. So think of
1:00:30this that in a company setting where
1:00:32many freshers join or everything how
1:00:34helpful this is because otherwise I have
1:00:36to go and understand the changes and now
1:00:39at least a vetted PR can come to me or
1:00:42to the higher management just making the
1:00:44work of a tech guy a lot easier right
1:00:48and even from a non-technical
1:00:49perspective this will tell a lot of
1:00:51things right so that is the whole idea
1:00:59great So yeah with that I really hope
1:01:01that everyone we have understood the
1:01:02same and you can easily just get this
1:01:04workflow downloaded again if you want
1:01:06the same has already been uploaded
1:01:07though so you can use it directly just
1:01:08make sure that you are able to fix the
1:01:10error the way I've told you in this
1:01:12particular video
1:01:16great everyone so let me know if you
1:01:17have any doubts anything which is not
1:01:18clear we'll be more than happy to help
1:01:20you in Yeah,
1:01:30and I hope you like the project right. I
1:01:32think the project is really helpful as
1:01:34I've shown you by showing the thing
1:01:36live. So let me know if anyone has any
1:01:39doubt. Private tutor I want to create a
1:01:41private tutor that helps student to
1:01:42solve their problems. So uh Pinku what
1:01:44we can do there that's a very good use
1:01:47case. uh we can create that right we can
1:01:49create a private tutor which can talk
1:01:51and converse with uh the student and can
1:01:54solve the problem as well. Okay
1:02:02bro, can you create it is possible to
1:02:04get an AI agent which can tell you how
1:02:05much views likes the content will get. I
1:02:07don't think that will be very valid. It
1:02:09you you can just give all the YouTube
1:02:11data and that will be helpful. Okay.
1:02:14Can you make an editor which uses nano
1:02:16banana? There's an option of adding
1:02:17image of a model for consistent face and
1:02:19option for dress and pose. Uh as I'm
1:02:20already working on that hopefully you
1:02:22will see the same uploaded on my channel
1:02:23tomorrow. So as I showed you, I have
1:02:26actually after spending lots of time, I
1:02:27have worked it out. Uh so there are
1:02:29again different different ways through
1:02:30which you can use nanom banana right.
1:02:33This is the one which uh normally you
1:02:36would see. So let me go to Google and
1:02:39from there I can show you nano banana
1:02:42API.
1:02:45So either you can use it directly or you
1:02:47can use via this file.ai. Now via this
1:02:49file.ai it is difficult because it asks
1:02:51you for credits. But don't worry, I will
1:02:53be creating the workflow. So I have
1:02:54actually already created one where you
1:02:56can see full workflow with image without
1:02:58image. I'm just testing out the same and
1:03:01yeah then it will be pretty good. So you
1:03:04can just upload an image and you want
1:03:06you can just tell what you want and
1:03:08finally we will be able to see that. So
1:03:10let me just show you an quick example.
1:03:14Let's say change the logo to Amazon.
1:03:18make me
1:03:20wear a jacket
1:03:23and add lightning. So, let me submit
1:03:26this.
1:03:29Now, again, it will take your image,
1:03:31change the image. All these things will
1:03:33be happening. But let's keep on to the
1:03:35today's use case, right? The today's use
1:03:37case was also a lot helpful. So, now see
1:03:40the same is happening.
1:03:43So, the major challenge was in getting
1:03:44the API and everything, right? That was
1:03:46the major challenge here.
1:03:49And once it gets the file, I will show
1:03:51you that what it has done. Okay, so
1:03:53let's see the result view.
1:03:57Okay, not sure. Okay, I I I meant the
1:04:00other kind of lightning. But it has done
1:04:02some other kind of lightning. But I like
1:04:04this thing very nicely. You see how
1:04:05nicely it has integrated the jacket,
1:04:07right? Even if someone showed me this
1:04:09picture, I will also get confused
1:04:11because I have this kind of a jacket.
1:04:13That's another problem. I have this
1:04:14similar kind of a Levis jacket. So this
1:04:17is what I like. This Amazon is what I
1:04:19like. This lightning it went a little
1:04:21bit overboard. Okay. So I will be
1:04:24showing you this workflow as well. Don't
1:04:25worry.
1:04:28Cool. Let me see the questions. Can we
1:04:30add another agent on top of this? Yes.
1:04:31Uh Raja, we can add any kind of agents.
1:04:34We can at least we can have a multi-
1:04:35aent system. So that all can be done.
1:04:39How accurate is it since it's AI? How
1:04:41much can we trusted? So Shaman again
1:04:43should be fine. Like if you just
1:04:44properly prompt the same it will be very
1:04:46easy and you can trust the same. Okay
1:04:48helpful session. Thanks a lot shdevi. Uh
1:04:51okay man manual intervention will be
1:04:53required I guess even though it is
1:04:55automated. So shan again if you will see
1:04:56this in action right it is not in any
1:04:59case it is doing something it is just
1:05:01getting the comments. So if even if the
1:05:03comment is not clear to you you can
1:05:05surely go to and go to your tech lead or
1:05:08tech person or the senior and guess that
1:05:09okay lead generator for exporting Indian
1:05:12handiccraft product. So push plata that
1:05:13can also be done. Uh I will be doing
1:05:15that end to end in either one of our
1:05:18courses. So the course which we are
1:05:19doing everyone right and you can use may
1:05:21as a coupon code it will give you 15%
1:05:23off right. So for people who are staying
1:05:27till the end we give this coupon. So
1:05:29that use may for 15% off on course
1:05:38right
1:05:45so I think this will be helpful for all
1:05:48of you
1:05:50let me mean I see more comments so
1:05:52thanks a lot for staying uh I will just
1:05:53be taking five seven more questions and
1:05:56then we will close it okay
1:05:58Please ID on AI tutor. So Pinku I will
1:06:01be uh creating that and I will upload
1:06:02the same. Don't worry. Okay. So we are
1:06:04slowly slowly moving forward with that.
1:06:07Can you explain what is an AI agent? Uh
1:06:09okay I've done that multiple times. What
1:06:10I will suggest for people who are not
1:06:12aware about it and in any case what I
1:06:15will highly say is just go uh you can go
1:06:18to my channel or even Krisha channel as
1:06:20well where we have uh explained about
1:06:22this AI agent and what all is this. Ply
1:06:26what you can do I have a master class
1:06:28specifically focused on NA10 right it's
1:06:30a 7 and a half hour class you can just
1:06:33see this right it has 122k views 4.5k
1:06:36likes 180 comments so doing pretty well
1:06:39right so you can see this class where
1:06:41I've explained each and everything in
1:06:42depth so about automation agents what's
1:06:45exactly they are and the use case as
1:06:47well right we have created so many
1:06:49different different workflows here
1:06:50integrated the same with telegram so all
1:06:52that thing has been done here Okay.
1:06:56So I will share this with all of you so
1:06:58that you can go via this
1:07:04which course are you talking about? Uh I
1:07:06think the link has been added any so you
1:07:08can go through that. Okay. Are you using
1:07:10a paid version of NAN? Uh techos I don't
1:07:12use the paid version of N. I have the N
1:07:14for my clients and the way I do this
1:07:16consulting I have the N hosted on the
1:07:18BPS. Okay. So on a virtual machine we
1:07:20have that. Just halfway through your N
1:07:22masterclass video. Yes. So that will be
1:07:23very helpful.
1:07:25Uh which course are you talking about
1:07:27that is done? Uh then good one which no
1:07:29code AI tool can be used for developing
1:07:31a website from scratch. Uh again Krishna
1:07:34you are getting confused. Website is a
1:07:36different thing. Website you can use
1:07:37maybe lovable or some drag and drop kind
1:07:39of a tool that is front end. This is for
1:07:41workflows and for the back end majorly.
1:07:43So any is not giving you a website
1:07:44right? So for that you will have to use
1:07:46something else.
1:07:48So please your seven hours video dubbing
1:07:50in Hindi. dubbing to pink I'm not sure
1:07:52but I can create one in Hindi if you
1:07:54want okay dubbing doesn't give that fun
1:07:56please everyone let me know if any doubt
1:07:58is there with respect to this workflow
1:07:59with respect to anything uh first of all
1:08:01thanks a lot for the person who created
1:08:03this but since we were facing this error
1:08:05here so I made sure that I can correct
1:08:06the error as well okay so now with the
1:08:09error corrected which was there in the
1:08:11official workflow I hope all of you can
1:08:13see that how nicely we were able to get
1:08:14things done okay and yeah I think we
1:08:17will we still are using this and we will
1:08:19integrate this into more of our use
1:08:21cases.
1:08:26Hello sir.
1:08:28>> Yeah. Okay guys, so do definitely check
1:08:31out the generative AI for everyone
1:08:33course uh including professionals and
1:08:35leaders. Everything is given in the
1:08:37description of this particular video. Uh
1:08:39we are currently running 10 to 20% off
1:08:42you know and I think there are only 150
1:08:44seeds that are remaining. Uh the batch
1:08:47will basically start from 14th of
1:08:49September. All the information is given
1:08:51in the description of this particular
1:08:53video. So yeah, this course can be
1:08:55joined enrolled by non tech anyone. Uh
1:08:58you don't need to have any coding
1:08:59knowledge because there we are not going
1:09:01to focus much on coding. Okay.
1:09:04>> Perfect. I think uh we can stop the
1:09:07session. Uh
1:09:09>> cool then. Yeah. Thanks a lot everyone.
1:09:11Just reach out to me if you have any
1:09:12problem and I think I will be more than
1:09:14happy. So I'm just sharing my uh
1:09:16LinkedIn and YouTube quickly so that you
1:09:18can easily reach out to me. Okay. If
1:09:20anything is there.
1:09:22Yes sir. I think we can close the
1:09:23session. It
1:09:24>> Yes.
1:09:25>> We were able to
1:09:26>> Okay, guys. Thank you.
1:09:28>> Find out all the information in the
1:09:29description of this particular video.
1:09:31Thank you. Bye-bye. Take care. Thank
1:09:32you.