Hello everyone, I am Swati and welcome to the Complete Crash Course on AI Agents. Today The class that is going to take place is very It is going to be important because AI Agents or Agentic AI Today's Date The most powerful, strong and If there is any revolutionary technology then it is this Is. Now all the companies that are there today Whether it is a service based company or Every company should be a product based company AI in products and services It is integrating. So far we have used generative I had heard about AI. In generative AI There are many things that we do day to day in today's date I use it in life. Whether it's yours Google's Gemini or Chat GPT As if we were models, we use them day to day I use it at work. that is all about the Generative AI. but also from generative a.i. A step forward and a next level technology which That's the AI Agents. whose Everything about this in this complete crash course We are going to read things in great detail Are. Now what will be the benefit of this whole thing What after completing the crash course What advantages are you going to get? So the most It is very important for us to I need a particular job. because without a job Does it make any sense that we should just keep reading? It doesn't happen. We have to reach the job. So we are upskilling ourselves. Now look for jobs in the market and only then create them It happens when companies have so much amount It happens, so many hirings take place. And AI in the market today The investment is more than $200 billion It has already become too much. That means every company Wants to become an AI based company today Is. Due to which the market explosion The demand for these jobs is 2.3 India has demand for millions of professionals In. But there are many people who I don't have the proper skill set On the basis of which you can get those jobs. So That's why we are so worried about the crash of this AI agent Introducing the course to you so that after reading this, whoever you are, AI driven questions are asked in companies You can crack the interview in them Got it. Now the most important thing in today's date is those are the things every company as I said An AI based company is going to be formed. That means every SaaS company means Software as a Service The company will need AI agents It is of this. So AI agents are not like this There is technology. No matter what tech you use Go for the interview. whether you Software Developer, DevOps Engineer, UIUX, Data Engineer, Data Scientist, Product Analyst, Data Analyst, Quality Analyst Engineer, Business Analyst, HR & Talent Acquisition, Social Media, Sales, Accounts, Partnership, operations in any field Be it a tech or non-tech job role. every Increasing the productivity of company workers Excessive use of AI agents for is going to do it. And this is the reason that very Agentic AI is the focus of all interviews Questions are being asked. specially for Machine Learning, Data Science, Data Analyst & AI Engineer. so here it is When questions are being asked I If I share my experience recently. I was there in a part of a panel interview Where people don't even know. AI tools are being asked in interviews And what is the difference between AI agents? No one is able to tell. Someone speaks that agent ai agents are the same as ai tools Is. So many people don’t even know the basics I know it as of today's date. In the interview Recently a question was asked. One It was a very good startup based in Bangalore Hiring was going on there. one of that The interview question asked whether LLMs are related to the internet are accessible in? Now, so basic People don't even know the answer to the question Was. That's why I am going to bring to you guys an AI agent This special course will prepare you for a crash course I have been. Inside which you will find all From basic beginner level to advanced level Whatever things are there will be covered here. Now Here we consider every thing as basic We will start from the foundation level. If You don't even know about AI today What is an agent? How does an LLM work? Is? So we will look into every single thing very minutely and We will understand with real time examples. this which The crash course is going to be of 1 hour This is going to be a crash course on around. If you go to give interview of any level In today's date or if you are a college student Be it a student or a working professional or You are a fresher, how can you afford this 1 hour lecture? It is important to attend anyway because Even when I sit for hiring On your platform, you have many people I don't even know the basics. so I Here is such a course for you people I am preparing you for any Go for interview, whether you want software For development or for data science. This course will be very useful for every job. It is going to be more helpful. now here These are the notes of this course I have written all the things in the description box below I have kept it on my wallet so that it's always with you It could be available. Now in this course, as many We are going to cover all these topics We cover all the important topics of do. How do agents work? LLM Chat GPT Architecture Agentic AI Mans AI Single Agent So as many as Important topics you are seeing all these I'm going to cover something and this class There will not be only theoretical classes that you Only the theory has been explained. which you Popular AI agents and how to use them If you want to know every detail about them then We do everything practically in real time Those who understand through examples Are. So let's start with this crash of course and we will do it in full place manner People complete it. now in some people's minds This question might be coming to my mind that if we do LLM then we Know about Large Language Models such as We used Chat, GPT, Gemini, etc. Is. Now when this was there then it was about AI agents what was the need and these LL.M. What was lacking in it due to which we Had to reach out to AI agents. So here it is The first level which came was of LLM Everyone will remember that when the new chat GPT was introduced When it came, it was very shocking for us that hey four gpt can even write the code Ess writes down whatever we ask If he answers it, then this was the level It was the beginning of the forest where we Looked at a lot of things in generative AI Whether it is chat GPT or image creation Be it mid journey, daily or videos All these things should be the source of Open AI where a model would help us generate things I used to do it and give it. After that we will move towards more AI went to the new next level where AI work People started talking about Flos. what is ai Work Floss? Here we are in a little while I will understand. Then we moved to the next level. Which are AI agents. Now when any one As technology develops, slowly that time It continues to evolve with the. there was a time When we use thick phones with very large keys I used to use it. but as of today We have slim touch screen smartphones It is available. Right? So technology It evolved gradually. in the same way Started out with LLMs. of A.I. A new era began. then we work floss moved towards and today what we are We have reached the level, that is the AI Agents. So slowly, one by one, we You will understand the level. Only then will you understand this It will come that AI agents are so capable today Why is it more important? So first of all no We talk about LLMs. So this There are many popular LLMs available today. You guys might also be using it during date. Four Everybody is doing GPT. It is meta, There is Gemini, there is cloud. there are so many such LLMs are available. Now this LLM work How to do it? What does it mean? So we go for LLM. Any input We give it. Like for example if I I want to get a let's suppose email written. so I I'll move the chat to GPT. to him I will tell you what is inside the email Need things. Then after that in LLM everything will be processed and what do I have to do will get? A plain output will be obtained. But Here are the limitations of LLM models That is the limitation that he has with us There is a lot of personal information There isn't much access. like I right now Let me show you this practically using the chart Inside GPT. Allright. so here is the Chart GPT and here I am just writing A Prot. let's suppose I am dating someone The meeting has to be conducted. For him I am a I am getting the mail scheduled. so i have to right here i want Two Schedule a meeting with Mr. Lion Manager at AYS sessions right here I I gave you the hint and here I am talking I am here to write a mail for me ok so what does chat gpt do here I get the proper presentation which I have I have given it to you according to which I have generated mail He will do it and give it to me. so here we can see If you have any email then write it here I generated the chart and gave it in GPT. So Obviously LLM is a generative AI based There is an application. what you wanted is here It has been generated and given. but now i'm here Pay Chat GPT Se Next Prop Agar Likhti Am. when will I have a next meeting With Mr Singh? Please let me know. So What is the response of chat gpt to me here Used to be? I don't have access to your Calendar and Mr Singh Schedule. so here But he is saying that I have no access It is not a calendar. so this is it here There is a limitation that the chart which is in GPT is based on the training data it had Accordingly it is responding. But It has no access to my proper calendar There is no access. because of which the question I am asking him this, I want his answer I am not able to do it. So here we have this I suggest you send your email yourself Check it. Check the calendar or a meeting If you have notes then look at them. Right? So The Thing is that here near LL.M. which training data is as much as it knows It will respond just that much. But if I tell him Beyond the training data, no personal information Or if I ask you some things then in that case he would not be applicable and he would have no objection regarding the same. There will be no access. so here we have seen that That's what Chart GPT is replying to us that he doesn't have access to our calendar that is to schedule the meeting Got it. Right? So what is meant to say here is What are the limitations of the LLM model that is, despite being a trend on the large amount of data even though we With any large amount of data train for LLM but they have We have limited knowledge about our personal of information or any company's internal Data is his right so LLMs are passive that is, when I get it, when I write I tell him what he does then Responding starts here Is. So whose limitations are these? LLM that the models should go and see our calendar Cannot access directly. He is my one Cannot work as an assistant. to whom I can say that as many inboxes as I have I have received an email, go and check it out And reply to those emails. This work is a Cannot do LLM. this is its limitation Is. The job of an LLM is to give him a It is trained on large data and The more information he has, the more Accordingly it gives us the output. i.e that this is the application of generative AI that if I will send him a message and he will respond And what will he do for what I want? He will generate it and give it to me. but my If you want to complete any task on my behalf Even if anyone wants, this is not a possibility Also near LLM. no matter how much he Also trains with large amount of data Let's do. So, we proceeded here And the next step is where we reach That's called AI work flow. Now what How does this AI work flow? That means that my Whatever my work flow is There should be someone to manage it. So here I am asking the same question that when will I have the next meeting with Mr. Singh. So this is what happens in AI workflows. How will things happen? that here we have put what did you give? On input we asked this question I told him that my next meeting is Mr. When is she going to be with Leo? it We gave it on input. Now this is the input Where did it go? Went for LLM. Now LLM The number of tools inside AI work flow If there is, he can access it. i.e that my Google Calendar is called LLM will access and after accessing my will check the calendar and after checking I will give me the final output that your The meeting with Mr. Singh is tomorrow It is 10:00 o'clock. So after LLM what we The next level we took was AI work flows. Where LLM has a capability that he directly gives us the tools What can he do if he is there? Access Tax Will get it. So who has the LLM here? Had access? Had access to Google Calendar. he took particular access and told me This was the final time of the meeting on the output I told you. So the AI work flow is one step beyond that Moved ahead. Now here if I pursue this LLM should I start asking you that the day when my There is a meeting scheduled with Mr Singh What will the weather be like that day? Right? so here But we gave these things as input. This is the input Went for LLM. But now after this LLM Has the pass limitation arrived? possess this LLM So the only access I had was The tool of Google Calendar was just for this There was access. These are the weather ones that we have are there APIs or from the link we had Was not connected. that's why we here The output that will show us anything here Pay output will not show properly. so to say This means that when we look at AI work flows, If we talk about it then there are LLMs in it he has proper access of the tool. Tools he will have access to He will answer accordingly. if his If you have access to Google Calendar It will go and check the calendar and in the output It will tell you the time. But now this LLM If there is any API related to weather near it If it is not connected then in that case it will tell us Any output that is there is in a proper way The show won't be able to. So explain what you mean What I mean is that when we do LLM If you start giving access to the tools then this Things happen inside the AI workflow. Now whatever I am doing in this AI work flow whatever tool I want to give it access to Or whichever API it is used for, according to it Will Work. Now let's suppose here that We have LLM, it has my calendar I have access as well. it can access There is also a weather app. And I call this I am pursuing this LLM so that when you get the time Confirm when with my Mr.Singh There is a meeting, how will the weather be on that day? and beans What I need is in text format I don't want it. I am in audio format I am going. so here I have to tell you particular LLM will tell you that you are having a meeting with mr singh at 10 am and the The weather for that day will be cloudy. Right? So I want this thing in audio format Am. So within a work flow we have multiple You can give access to the tools available here Are. If we want this, we can do it by text You can convert it into audio. so which You are giving this particular thing here Who is deciding? the output is correct Are getting. But here, however much you This can give access to multiple tools To LL.M. But here, as many paths as Like for example this calendar will access it. Who is deciding this? I am deciding because I LLM has been given access to this. That means that Who can control this path here? Used to be? This path at last is controlling is the human decision maker. That means that I I am deciding which tool to use Access is scheduled to take place next to the LLM. This machine is not thinking on its own. Machine She is not deciding on her own. of the machine I didn't have that much capability when I talked to him I was asking as if I could make charts from GPT I was asking. so chat gpt is mine Not connected to my Google Calendar. That's why they're telling me that I have There is no access to your calendar. My Don't have data. This is how AI works I will decide whatever the flow is LLM has to give me access to it or No. So what I mean to say is that as much as The things that are being decided here are Human is the decision maker. There is no AI here does not have the capability to defend himself Think about it. Connect to the calendar manually do it. these things are not possible here It has been happening. So what happened here? what did we Things seen inside AI work flow It was better that LLM, at least we got connected with tools but which LLM with which tool It's still human The decision maker is doing it, that is, this is the use of A.I. There's just one work flow. Now here the we talk about the next level that is About the AI agent, that is, an agent like for For example, we have never seen anyone in our lives This is not the first time I have heard the word Agent Is. We have understood the meaning of Multiple Times Agent I have heard it many times in many contexts. Like for example if I get a ticket from IRCTC I have to book my ticket somewhere. for example I want to book a ticket to Bor. Right? So now such situation happens here many times I feel like I have to go somewhere very urgently and tickets need to be booked immediately. So What do we do many times? to the agent They say that brother, please come on our behalf it is only me but on my behalf you don't need me Book a confirmed ticket for the bore. So now this What happened in the case? The agent here is What is his job? We have done this in this particular Gave a goal to the agent. What is a goal? That I want a confirmed ticket in my hand I am getting bored. Now the agent on my behalf what will he do? Search by going in different routes Will decide which train gets bored. Right? After that you will see my priority here that I got a ticket for second class AC Needed He will see. different for me now At the junction like for example that I I am getting bored with Delhi. So from Delhi Which routes will be there between the bores? which route I will reach there soon. I get a confirmed ticket In which particular train will it be available? it All of this work, all of this reasoning, Applying logic and having the final ticket in my hand Bring it and hand it over. This is a on my behalf The agent does it and in return I give him I pay for the service. Right? this we Hear the meaning of agent in normal language Is. Many times we have consulted financial advisor LIC agents should first go to the He used to come to our house. Right? that our In front of us they used to tell us that this is the policy. If you do this then you will get this much return. So In such different segments, we have Have you heard about agents before? In common language. But here the AI agent This means that here an AI is a goal or a task will be given and that complete that task on our behalf Will do it. Now in AI work flow also we have I got to know about the limitations, brother, here too It is the human being who is deciding. But A.I. Such things will not happen in the agent. here AI self-reasoning capabilities will keep it. Now for example here is one I am looking for a task for AE Agent that Create an Instagram and Lindin post. Post an Instagram for me and Lindin I want to make it based on the news articles. That The recent news that is going on According to me Instagram and Lindin You have to write an article or a post Based on news articles. and every day to him To be published at 8:00. Right? this task Who am I giving it to? An AI agent The first step in this will be The best way to do the right reasoning of the region is to what will he do first like for example if This job would have been given to a human like us So what do we do? First of all, we will win Also see the links to news articles After reading each article, Read four articles and then one article We summarize and then write a post. Let's start it off in the same way here's a The agent also fulfills the task given to him. This It will be divided into different steps. Step It will compile all the links it has Will do it. After that it will summarize the article And after that I will write the post. Right? so here But this task got divided so that I can go first I will do this, I will do this, then I will do this. Right? So in the same way, an AI agent has decided this. this in step one Have to do it. This is in step two, this is in step three These. But in step one when I compile If I want to share links then which tool do I need? would need? the way we humans think Right, hey, should I use that word? No yaar in word then link properly It will not be possible to align. Let's think again Should I use Excel? but excel is Good for managing data. Let us go In this case it will be best to use Spreadsheet. This is how we follow human logic Let's apply it. In the same way, AI agents He will have so much capability that he can handle all these be able to reason about the questions i.e. Be able to build logic. Now this is for reasoning firstly he will think that the best way What is compiling news articles Of. Can I put everything in a Word document? should I copy paste it? Now all these After reasoning he comes to a final conclusion It will reach that this is compiling the link This is a better option. Now this we This seems more sensible to me. Let us go Now I started looking for tools for this. Am. Right? What will he do now? Finding Tools For this, separate reasoning will be applied again. this now Look here, what was the first thing? He of reasoning. he thought to himself. Right? As We humans think in the same way as they think Applied. Now after that whatever reasoning is to be done After that he also needs tools to do the task To complete it. So he thought he would say the word What? then I felt the word was not better Is. Then he reasoned that Excel is also good Not there. then he thought that yes spreadsheet It's the best. Then summarize the article To do this he used Fi-To. Now this It is not even better. Then he took a separate Used LLM. He doesn't even have that better Put. Then after that he finally felt that Perplexity is better for this job. He I chose that one. then to write the post for he called the cloud of Claude Anthropic I used it. So this is like our human reasoning We do. Correct? This is how he reasoned here of. and here is a please a framework It is used. This is called react i.e. RE of reasoning and ACT of action. Right? So what will he do here? Reasoning first Will do it. then for every problem What tool will I use? it here The action will be performed by We Are Tools. So this Doing reasoning here one by one method, What tools do I need to use? These He will make the entire strategy. His After that he will finalize that yes is the first thing have to compile have to summarize the links Have to post more. so I'll post the link here Sorry I used excel sheet to compile I will use a spreadsheet. After that Best for summarizing articles and AI cloud to write the post. So this How many before reaching the final conclusion so many iterations, so many decisions, so many all the logical ability, who knows all these things doing? all these things here It is not happening through human intervention. AI can do so many reasonings on its own performing and then going to the finals reaching the conclusion that yes, everything works Which one have I found suitable for me? Is? The tool has been received. What will he do now? This Access to every tool AI has to offer has the capability of. access to those tools Will do it. and my final output is what is the output that I get here You have to write and post your Lindin and Ina Pay. Who will do this work also? Ai agent will do it. because it has accessibility can my Instagram and Lindin's Too. Now look at AI work flow and your What is the difference between agentic AI? now here If we had an AI work flow So so many steps, so much brain was also used I would have taken it and written the post. But if let's Suppose I want the post to be funny. Right? So Give me iterations here again and again Would have fallen. I let's suppose chat gpt we guys I use it. He emailed us once He gives it in writing. so it looks like this hey So it is a very machine oriented language. This I want to make it a little human oriented. So We rewrite the chat to GPT That makes this sound more like a human. They He writes it again and gives it. Still good to us I don't think so. We want to add something else to it want. I want to make a funny post I will give iteration again. So A.I. Who does all these tasks in the work flow? Who are the iterations by giving these prompts again and again Does he get it done? A human gets this done. so this AI workflows have a limitation No matter how many iterations are performed who does that? A human does. Right? Now the same thing if you have an A.I. If there is a work flow then a human is doing it. But you will have an AI agent, so how? I will tell you how to perform trishans Real Time is an AI agent which is from China The company has developed a very best AI agent has prepared a medicine named Mans AI And when this manslaughter came in the whole market There was a stir because these worlds The first autonomous AI agent. how does this work does it? So I'll tell you about real life Demonstrate practical things here too I am the one. So let's move mans ai By visiting the agent's website and using it Let's see. So you have this thing about iteration You will understand it right now. so let's move Into the website of man AI. and here it Is this is the website of Man AI and here In this Manas AI website I have seen a Proper praise has been given to it. i want to Plan a three day trip at Delhi. please Plan out the trip for me. End Accordingly create a website and dogs For me for this trip. Right? so what am I I want you here? here is one I have given a particular task that I will do three I want a day trip to Delhi. His You can plan a trip on my behalf Do it and I will also create a website I want it to be related to the trip If you have all the information then you can also create a website from me Create it and give it to me Regarding everything related to this trip ination is of him to me in a dogs format You also prepare all the things and give them to me. So Here is one thing which I am doing in particular I want to get this done by this AI agent. so here But see what is special about it? Manes said what did you do? now first of all for this work Which are the good places to visit in Delhi? is it c? For this, it first needs data Will have to collect it. so here it is Speaking of connected to data source. Everyone from Delhi is doing research There are attractive places about it. Here we will see from different places that Different commands that are being executed Are. After that the file is being created here Is. Searching, browsing. So what is all this information doing? Look at all the websites here too You have been given a link so that as soon as you You will open this website link automatically Here I will show you who It has taken data from Wikipedia. India Gate data is from Wikipedia. This is data from Delhi. So whatever things are What is it doing with all this data here? What do we need here by collecting it yourself doing? is doing the show. Right? so that is About that. Now after this when it has processed the data He gathered everything and then he separated them Start syncing things here so what do we need here? We He was told to create a website as well You have to give it. So all this information When you use it yourself once, All the iterations that are happening here. Look, I had some intervention here. No. Right? It did everything on its own was generated. Finalizing, Trip Make a planner. then after that here Which command to execute. So Till it gets the best result, we People can see it here. If you start from the beginning I looked here, on this line, I had given a prompt. and if I put this down If I keep scrolling down then as much Browsing has been done, searching has been done, files these iterations are created automatically You have performed again and again. without any All these problems can be solved by human intervention on its own. What have you done to things? created it Is. And what is the final result we got? Man's Haze is written here Completed the current task. after that here If you look at this segment here He destroyed all the files has created for us I will do this as soon as I will go to View All the Files and click on it. Look at all the files it has created. If docs have been created then which ones? There are dogs, its folder is ready and ready. Images There is no one here. See the code files. Even here we talked about making a website. So All the files related to coding It also came here after being made. What links You can find the ones he used here. So Now the first thing we'll look at is We wanted to make this website Has he created it? So here we You will get to see that Delhi Trip Website It also automatically created a folder by the name Gave. I did not say that I had any I did not give him the command that you repeatedly do iterations or perform and website Make it. It did all these iterations on its own Have performed. that's why we say here Could be LLF level one. take level two in we looked at the AI work flow and AI agents are several steps ahead of that because It has created the website automatically. Now If we want to see what kind of website is made Simply go here and see the download The option will come. I can simply go and download it I can do it too. So here our The download has started. Now let's see By opening the folder we will see what We are getting results from the website which Bains has created it and given it. so here we How can I have this website by name The folder has been created and arrived. and different things Attractions, Guides, Indexes for For all this I made an HTML page here I have already given it. like for example I I am clicking on attractions here. And This is HTML which creates our website It has been given to us here, it should open here Used to be. Now the most important thing here is that There are other use cases that you can do with this agent I am going to tell you something related to it also. So here we can see that Delhi Explorer, this is our website here Pay has been opened. In the historic site As many places as Red Port, Qutub Minar, Humayun, Stumm is giving you all this information that particular look is the most important thing If a tourist goes somewhere then the place which also going to places like for example redfort If he wants then what is his fee? Timing What time will it be open? in which location Is it Redfort? So every thing here is proper I have given you the location and the timing Is. The entry fee will be ₹35. These Giving you all the information in one place By creating a proper website for you. Right? And as soon as we scroll this Which market places will come down? Is it good. Where can you do shopping? We can see a separate one of it here People have a different kind of tiles here It is made here. In addition, all your Information Museum Cultural Institution Roaming. Three Day Eatery Trip That Will Help You I have planned it out for you, you can also see it here You can open it and see that you have day number one Where should I visit? In day number two Where should one visit? where in the afternoon Should go. Where should one go in the evening? so everything the plan has been out prepared for me in the format of a Website. So now I can go anywhere You just need to have this website. So yourself with these few seconds for me my The website was also made and my entire Also became a planner. So what did the agent do? The task which I gave was completed 100% fully Completed my task in this manner. Now here I have a website. If my If anyone needs a travel guide then I am a travel guide As soon as I do it here, look at it I love it from October to March, April to June The entire travel guide here is mine She is coming close to me. where are you walking The distance is where to go by car, Where is the rickshaw, where is the taxi, everything Right? so it is the thing that we did the task based now this ai agent is The objective is that whatever you have Task: This work can be done by only one human I create a separate website for all these things An agent is able to create and give it to you. But this does not mean that if the agent we have our software Developers will lose their jobs or data No one will hire a scientist. These There are so many companies, especially AI Machine Learning Engineer at Segment, Data Scientists, developers, all of them will be needed. But any particular tech who is an employee You have to terx his productivity So for that, you have AI agents They are the most beneficial. So Ultimately we need to give people as many things as we can They are learning here, that is why they are doing it That our productivity should be 10x. We this level coder programmer that our If productivity is 10x then any company Be ready to hire us. So this All the iterations I was telling you about It performed all the iterations itself and All these things happen to us within few seconds By creating it, it has been easily available here. So this is a website that has been created here. For us. So besides this there is more to it There are some use cases. like for example I want to buy two stocks. Right? One One is let's suppose Nvidia's stock and the other is Let's assume the stock is Tesla's. Which one Stock do I buy? if I chart this thing If I go and ask GPT then the chart will be displayed on GT. What will he do to me? Write the induction of both that he will just give it. will not be able to take the decision that yes buy SPI Tesla or Haswati Buy Nvidia. not by deciding this Will he be able to tell? But this is the thing if we from manes or any other ai agent if we When people ask us, this is a proper decision I will tell you after making it because it has The power of logical thinking Iterations by itself Performing all these things is an AI agent It is available inside Is. So here we can see that now that A.I. When it comes to agents, the steps are the same Is. to compile, to summarize links I have to post more articles. Right? Now For this, now instead of a human, AI is used here. The work flow involved repeated iterations. The post has to be made funny. Keep the post polite Have to make it. The post needs to be made more humble. So There were human iterations. but here at Human's place like its previous place We saw all these human iterations in Were. But now there is another person here instead of a human The second LLM has arrived. Critic Bot is here Already happened. What is Joe doing? By myself Performing iterations. like just now We saw Delhi for 3 days in Manest wala trip for example automatically search that Was doing. I was creating the file. The number of iterations performed by itself Was doing. No one is going to interrupt me in this There was no need for it. so basically do this There was a critic bot. So here you can win also there were iterations happening in iterations as many as Questions were coming that we have to make sure that it be perfect. all requirements Whether it is getting fulfilled or not. I am pursuing this LLM to anyone reading this post from another LLM I will get it checked and will be repeating it till then Until the best criteria is found. So whatever reasoning is going on here was using his logical ability, he was who was doing what? Your LLM is performing all things on your behalf Had been. Which comes through an AI agent. After performing all the iterations manually What did we find here? Final post. Like we just saw in Manas Received the final results. create folder automatically Found it. The documents file has been created. so this Everything we wanted as a final output The task which we had to get completed. AI The agent performs that task on our behalf He completes it and provides it to us. Right? Is this clear till here? now after that Now this AI agent's work What is the flow like? If someone asks you First of all let's conclude that first of all The goal has to be defined. then this reasoning He performs. then for this reasoning What tools will help him? will perform the action. Perform Iterations and after that we will get the final which is ours Output is output particularly Provides us here. Right? Now What happens when AI agents? how it works does it? How does iteration perform? Till now we have understood everything Is. I am hoping that you will get the clarity will be. Even the number of people who are given the class Looks interesting, I understand it, Don't let me know in the comments. In the comments Let me know so I can find out how many people I am understanding the class. Moving Forward Are. Now this AI agent is the most Probably we are here in two different You can concur with the type. There happens to be one There is a single AI agent and there is a multimodal AI Agents. We have both of them here We will understand one by one. First of all we are here Let's understand what a single AI agent is It happens? How does it work? So here it is Like for example right now we are together here But we are taking the example that now here I want to plan a three day trip to Tokyo within budget. Right? so what will happen in this case What task did we give him? Task Here is a 3 day Tokyo trip plan Have to do it. Right? Now for that LLM which It will happen here when there is a single model here Then there is only one LLM. By single name It is now clear to me that we are talking About The One LL.M. so inside this agent There is an LLM which teaches different tools will access our answers and will respond. Right? So the task is for us It is clear. Now whatever answer I have I am expecting that from my model I have these expectations that I I need detailed locations, cost, Where is the hotel, are tickets available? These I need all the initiative from whom? This From LL.M. From whom do I want this answer? From your agent. So now let's suppose here First of all, who will prefer which I will have LLM which will give me these answers Could I provide it? So first of all this model will select it. like for example here The LLM which this model selected That is the anthropic cloud. Right? or someone and Preferred Model. So now here this LLMs are the demonstration of all these reasoning Will do it. So the first task is model selection Of. Now after this the task given to him If it is there then he will search Google map for it. Will take different sources. booking.com Credit card will be taken for subscription. all of these You can access the information here and Then we have to display the final results Will do it. Right? So this is the work of a single AI Agent's LLM is the only one for all your needs is performing the work and your There is a task and I am preparing for it. But When we talk about multimodal agents here If yes, then there is more than one LLM. are used inside the agent and they are our Let's perform the task. Absolutely simple If we talk about multimodal agents So an agent who has two models I will remain prepared. that is, two like for For example, I have a particular one here Do you want to write an article or a post? I need to get it written. Ok? so here if So what if we have multiple AI agents? will be? The one who is my agent is a writer Will Work. and the other agent will check that The first agent who is writing is writing correctly is he or is he not writing. so simply if we If people give example of this multi eye agent then Things happen with the two agents as well. Both agents have their own programming There are codes. But what are both of them doing? One Two agents to perform the same job I am working here. Right? what is the work Is? This is the work. The task is that I have a I want to create a good proper post. So One agent will write the post and the other will be the editor The agent will check it. That means both of them What is an agent for a single task? Link you are here. Right? So this was a simple Example of agent from. Now look at the two agents. Yes for example. So an agent like for For example, this is my writer Agent Blue and the green one is mine Editor agent who will check. So writer's agent What are you doing? working on your writing Is. But this agent number two is the first He will do his work also. I will keep editing. Check I will keep doing it. but the agent who is working he will get into his work also and I will keep checking again and again. so what I mean to say is This is when there are multiple AI agent models So in that an agent can do a task and The second agent can also perform a different task Or it could be Agent One who is performing the task. He can go and help him also. So in different criss cross manner here Each agent has his own task You can align the tasks. Now when someone If the task is aligned then it is obvious that Tools will also be needed. so like for Example here is Agent One. This Someone is using this tool for writing. Some task has to be performed by it. like for Example: now this editor is an agent. check this We have to make sure that whatever Agent One is writing is correct Is it or not. Is the grammar correct or not. so be it It could be any grammar based application using the tool. So what I mean to say is this Whenever we talk about multi AI agents If you do it, then there is more than one or two There could be agents. Both agents are different Can work as either or both agents We can do the same work together as well as separately You can also access and use the tools. He also uses his own tool for each task. You can access it here. so this There is a multi-model AI agent where one We use more than LL.M.s here We do. Now in this only when you do LLM If you use different agents then You can approach different agents in different ways. You can use it with different patterns. What are such patterns? on this topic In particular, you should pay attention because Many times the interview question becomes that If we had an agent to prepare us If you want to do it then follow these simple steps to perform this task. In which pattern will you use these agents for will you arrange it? In Sequential, Hybrid In, in hierarchical, in asynchronous. so this You should know the topic. So let us go now People understand the patterns in detail here About this. so now because multimodal If there are agents, there will be different agents. But What pattern should every agent follow And in which pattern is it kept Mostly people understand this. So the first one The pattern we are going to understand That is the sequential pattern. Now like for Example here is an AI powered Document processing make us here Is. Ok? The job of the agent would be that an A.I. It will process the powered document. Now sequential means by name as here It is understood that there is an agent here Now I know that some tasks have AI power If there is document processing then an agent can do some will perform the task then move that task forward He will forward it and then the next agent will work on it He will do it and then he will pass it on further Agent Three will work on him, then one by one The work on the sequence is done by each agent Everyone will do his own thing. like for example Our agent number one here is this Agent Blue What did he do? To all First, the text extraction work was done Now what should he do? finally one We have to create document processing here Pay. So any number of texts like Wikipedia is a source, I took something from it. Is there any news article. let's suppose There is Economic Times, take some source from there Took it. So from each different platform What is the agent doing? to the text Extracting. And as soon as its The work was completed, so what did he do? This The work was carried forward. Moved it forward. Now Who got this job? Back to Agent Two. what is agent two doing now that agent one All the text extracted so what will it do to that text? good From now on he will summarize. Agent One Texted extract it and send it to agent two Gave. Now Agent To whoever has the text he has come, he will check it and will take it carefully what will he do? It will do the work of summarizing. Now Agent Two has done his job. Text I summarized it. Now he took this work forward will give. Now this work went to Agent Three. What will Agent Three do now? As much action as to access the item extraction will start working for and will get the final This will work to create the post. now when If his work is done then whose final decision will it be Did you move it closer? Back to Agent Four. Now The work that Agent Four will have is that You have to go to the database and save it. So Sequential means one person did the work next Increased. Then the other one worked further Increased. Then the third one worked further Increased. So here we work on each sequence one by one It is happening. That's why when we do it this way This is how agents work If we keep this then it is called sequential pattern goes. this is not exactly like an assembly line It is like that. This is the assembly of a big car it occurs. so first a bot in it first what will he do? He will make the body of the car. Then He will put tyres in it. then when the whole car is If it goes then it will be coloured and the final will move forward Will send it. so just like the sequential which There is an agent here, an agent worked. Then Second, then third, then fourth. one by one The work progressed here in the sequence. So this was our sequential pattern. Now Let us understand the next pattern here. That is the hierarchical pattern. now what is this It happens? So for example now here What is a task? that write a report for Business Decision Making. You will find here a Create such a report so that the business A decision can be taken. So now that the hierarchical it happens like we talk What happens first inside companies? There are many different positions. Interns They happen at normal levels. then after that In this you will have full time employees. A Like let's suppose someone in the team. Then After that there will be a manager over all of them. Then That manager will have a senior manager. so this What is? There is a heraldic. Right? in the same Manner like it happens in colleges. First of all Yours is Let's Suppose HOD. then after that You have your faculties. then after that here There are students. So there is a hierarchy. In the same way we use hierarchical There is a pattern. Now here's one that we have The main agent is called the manager AI Agents. and then here it is all Agents work in different ways. So this manager AI agent is given a task I've found that you need to create a report. It is for business decision making. So this Took this task first. Now after this First of all what is Agent One will he do it? will monitor market trends To. Only then will he prepare the report. then after that What will Agent 2 do here? as many as It will analyse customer sentiments. What will Agent Three do? as much as Analyze performance metrics Will do it. So these three agents do their respective work did. After doing their work they come back Who did you send the work to? to the manager AI agent. Manager AI Agent is one of the three agents, Two and three. Check the work of all three agents Did. Now it felt that yes, now the work has reached here It's okay. Then what did he do? All Agent Four was dispatched. sub agent four kya will he do it? will aggregate the final insights and will make a final result and after that our What will happen to this task? complete Will be done. so what happens in the hierarchy That one of your main manager agents was there. Then he The work was divided between two agents. So in different ways here which is now hierarchical when it comes to that it is called hierarchical Patterns in which one of your managers, an AI agent It happens and after that all the agents here There are. All agents are doing their work Informed the manager. then to the manager If the job seemed right, what did he do? Sub Agent Four was asked to prepare the final report and then produce our final output Done. So in this way the next pattern What happens is called a hierarchical pattern. So Is this clear to you guys? Moving Forward Are. Now this third pattern Let us understand here. Now the third The type of pattern that occurs is called Hybrid Pattern. like for example Autonomous vehicles. Right? So now what's in it will be? There is a top level AI agent Now if there is a completely autonomous vehicle then it If the car is driving without a driver Plan out the route first It will have to be done. Create a traffic strategy It will have to be done. So this is the work that this top does Level AI agents do. now after this All of them are agents too. So this is the pattern You must be seeing this pattern here even if you look just look at this As far as I'm concerned, I made the pink box... What we saw before will look like He was like a hierarchical agent. Right? But If you look at this pattern now, here Pay sequential is also involved. So Both sequential and hierarchical in which gets involved which we call Hybrid Pattern. now after this there are many All of them are agents. like for example There will be real time sensors to ensure that if the vehicle is moving If it remains then there should not be any collision in future. Ahead If a tree comes, a stone comes If it is there then who will make sense of it? So one by one The agents will continue doing their work here. An agent looks at traffic strategy Is. One is seeing if there is a collision ahead It is not happening. One is seeing that Which route is the traffic taking? So now All these agents do their own work will go on and a feedback loop will continue which will Provide feedback to top level AI agents will keep doing it and according to that whatever top It is a level AI agent, what does it do with its things Will you keep doing it? I will keep planning it out. So this the patterns that occur in the manner that is Called as the hybrid pattern. Next Let's talk about patterns. that is about What? That is about the parallel pattern that An agent did his job and then moved on. Then after that another agent came. So one by one Agents such as for example in large scale If you want to do some data analysis then The agent is doing one step, then another third. So in this parallel segment, if we People arrange it. so that is Called as a Parallel Agent Design System. After this comes another important thing which It is called asynchronous. like for example AI Powered Cyber Security Threat Detection is what we're looking to make. it is in that case This is for when there is a particular threat or No. Talking about cyber security Are. So now there are three different agents here. Are. What will one do? The network will be monitored will control the traffic. the other will see that someone The court is not able to see any suspicious pattern in between and the third one is random sampling will do it, will do test use case so that we can If any fraud is detected here So that you can inform about it. so this The task is basically this type of task that if some threat is being detected It seems like the asynchronous thing is visible This seems to be a very unusual activity Some agent is thinking that this According to that we have got the result It performs here so it is asynchronous That means everybody here does their own work is doing it and whoever feels that something Unnecessary conditions are coming, something different things are happening that didn't happen before then what will it do according to its response will start doing that is called as the Asynchronous I have come this far, hope it is clear to you people Let's move ahead. now let's talk here Types of AI agents. so here we are If people look normally, there are five types of AI There are agents. The first one happens to be Simple reflex agent. The second one happens Model Based Reflex Agent. would have been the third Is a goal based agent. Fourth is utility based agent. And fifth is learning based Agent. Each agent with example. Now Here, we understand. Let go Let's move ahead. So the first type of AI Agent is simple reflex agent. What is it does it? These agents are based on the current situation It works. if it meets the condition takes action. Otherwise it does nothing. No So it has no memory, so it There is a learning pattern. like for An example is a vacuum robot. as soon as If That's When It Hits the Wall Let's Suppose There's a wall here and you have a wall here The vacuum is a robot. Ok? so as soon as It will go here and will see the wall. These what will he do? perform a turn left will give. He will just take action. So if this This action only when the conditions are met and the ball is found will perform. Otherwise it will not do anything. So it neither has any memory nor There is a learning pattern. think of it is Like if this happens then I will take any action I will perform. Otherwise I will not do anything. Like for example sometimes we are like this doors see sensory doors that we have no When a person stands up the door opens automatically What happens? It opens. so here If someone detects him, there is someone here But it will be open. So if any action is He will take it only if he gets some conditions. So this is basically a simple thing in which There are sensors, there are action based rules and there is an actuator which activates the actions He performs. There is no memory Nor does it have any learning pattern. Now The second agent that we understand here that is about the model based Reflex agents. This agent has current and previous knowledge internally It uses the state. this little intelligent It happens and some memory is also used here It does this because it uses previous knowledge He keeps it somewhere. so that's why here It uses memory. like for example A Smart Vacuum Remembers Room Layout And adjust its route. Ok? so here what will he do? If there is an agent then he is particular He will see the obstacle. Now it's as if he has no obstacle What will he do if he sees it? memory that will last He will go into it and check himself and the map. If you have checked the map then it will follow your path will decide. Right? So here its They have so much ability that they can Performance, memory usage And according to that some action is taken here He takes the pay. There are sensors and they update remains with the internal state because it has Memory is used. if let's suppose An obstacle has come up. This went here. This When the obstacle comes, now identify the obstacle Done. So what will he do according to this? will go to the map. will look for another route and After that change your route here will take it. He will decide his direction. So This is what uses the memory which It is called model based reflex agent. Let's move ahead. Now let's understand the goal to the based agent. So these agents are goal oriented It happens. He thinks before every action. Will this action take me closer to the goal? Will you go? Like we humans think that lets Suppose if I have a particular goal that I want to increase my weight or I If we want to lose weight then we will take whatever action like for example let me wait if Suppose if I want to loose weight then I will avoid it Sugary things are extra calories I will avoid fat in my daily routine in my food so that only then my goal is to achieve weight So that I can reach it to loosen it. So In the same way, this agent is this goal It is oriented. perform every action I will think before doing this action am I doing this or am I doing this? Will you take it to the goal? so like for For example, if you have an AI agent to whom are you telling that my flight I want to book for Goa. but the flight The cost of the ticket should be around ₹5,000 It should be inside. So what is this agent now? will he do it? What task did he get? that to this Have to book a flight. what is the price segment Should it be? First of all under ₹5,000 This will go to path A. If it is sold for ₹ 6,000 If you get a flight, it will check whether it is Is the path taking me towards the goal? So the answer is no. because what is our budget Is? Of 5000. Then it will see path B. Path In B if it is 4800. is this action If he performs, will he get it near the goal? Will you take it? He will definitely take it. They This is as much reasoning in itself as any Before performing the action, he tells himself He performs. like we humans For example if I apply for a job I am going to give an interview. Right? So I will think about everything before going First of all, I wrote this line in my resume Is what is written correct or not? I am on my resume What should I write that will help me I was able to get the interview. Right? So every action the reasoning we apply before He has these goal based AI agents They have so many capabilities that this reasoning can be performed by itself We find it. Right? so that is about the goal Based Agent. So the next one is utility based Agent. This agent doesn't just stop at the goal Rather, the utility of every possible outcome, i.e. That calculates the success level. Best Chooses the possible result. like for example it choose the flight with the best Timing Price & Airline Rating. Who The base was round, we gave it a round shape so that it Book tickets worth ₹5000 inside. So he is someone Get flight tickets at any time He got the place, he went there and achieved his goal Focuses on satisfying. But This is utility based only Does not stop till the goal and the better achieves that goal What other outcomes can there be to do this It calculates that as well and best Provides the results. like for For example, if it is a utility based agent If it is possible then he will choose the flight. Which ones Flights are available at best timings, which one is best prying is pay and airlines because If they are different then which airline is the best? according to which one has higher rating He will choose it. So this will create option A, B, C According to the scores. And if option A, Utility of any option among B, C If it is more, he will choose her. It Ranks and choose the most useful option. So This is creating different options. Option A, B & C. And all of these Everyone has their own scores. The best of these You will have this option and they will recommend it to you It will show that it will only stop at the goal. Round So it is getting completed in option A also. In B It is happening too, it is happening in C too. But Which is the best among these three? in which May your goal get the best results. These to think about that task in that way Has the ability to perform. that is The High Level Utility Based Agent. Even I hope you guys like every type of The difference must be becoming clear. We started Where did you do it from? We did it with simple reflex It had started. simple reflex which is That was a very basic level and then it became advanced. Went to model based. Then went to goal based. But now utility based even beyond the goal We have agents available. so that is all About the Utility Based Agent. Next Agent Is About The Learning Agent. Now this How does the learning agent perform? Let's just understand that. Learning Agent. This agent tells you from his experiences Learns and improves itself with time He keeps improvising. like for Example: We humans do right, many times we People learn from their mistakes. not this Should have done it. So when ever such a life comes in the future If the situation repeats, then we should There is an estimate based on previous experience. right that you shouldn't have done this or done this It was needed and we took actions according to that Let's take it. in the same manner this agent also Do the thing. Learning will implement Performance. Then whatever problem arises You will learn new things from it and improve yourself A feedback will be given. like for Example Agent Find Tune With the Feedback And an AI assistant gets better at booking Trips for you. so an agent once he Booked tickets for you. to him a I had the experience. so next time his That task will come to you again. So that's in your memory will recall and more improvements will take it. the feedback he gets It will learn accordingly and it will be better for you The experiences that evolve you will give. so just like the humans do this type of AI agents have the same ability to think Like that. Because we use humans' brain It happens in neural networks. in the same manner When we create an LLM agent So these are neural networks. The one whom we read inside what? Deep Inside Learning Machine Learning. That is why it is which is your The Scale platform here The SQL for your Python machine learning All these lectures on Power BI are for you They are kept with me. You should go to YouTube once Will check out the channel. all these lectures You will get them. of die scale along with You get it for free on the website as well as the app Courses, Data Science, Data Analyst You will get it too. where the projects, Video lectures, notes, code files all You will get things in one place. still If you need any information then click here But the contact number given to you in this You can also reach out. So this which All five types of agents were reflex, Model based reflex then after learning Agent, Goal Based and Utility Agent in You must have understood the performance of everyone that How each agent works. so that is All about the types of AI agents. so how the AI Agents Work. any delicate task will come to this agent. Thereafter that As many tools as you need to perform the task Making API calls or accessing the internet do or interpret the code Reactivity environment saves your memory Access will perform autonomous actions Will do it. so this is basically how the ai Agents work. the next thing is that If we look at the scope and opportunities in this talking about so look every company that is There are different companies in different segments. Now our ultimate objective is that we People have to reach their jobs. Now to the job if we want to reach different there are companies where we basically We can talk about jobs. Health Care based companies. Right? For the finance There are fintech companies here. Retail Companies There are. After that, in e-commerce, The company comes here. Right? So In different segments or any product based company can be yours. No mother-in-law It could be a company that makes software. It could be a service based company. So There are different companies. Now every company has to Agents to increase productivity There is a need for it or it will be needed in the future. So work has to be done on these agents on a massive level will begin. Now like for example The segment which is related to customer care is always there in every inside the company that you let's suppose that I bought a product right as if someone There is a particular company in health care From which I took some policy, I I took their product now there is no If there is a problem, we will file customer complaints then to roll all those queries Agents can be extremely helpful that will resolve your customer queries Or let's suppose if there is a coder then If you want to complete any particular task then Help him as a coding assistant So in every segment, no matter what type Job ho apka agents every company introduce will do it. Why? because yours them You have to increase your productivity Of the employees. So for this every company Agents will introduce. Now different There may be agents of the segment. As If you want to automate any task. Like for example you are a founder Are you busy or are you a freelancer? near you There are a lot of mails. You need to send those mails There is need to check, have meetings and do research. So what will the agent do on your behalf? Your will check the mails. every important mail He will summarize and tell you. Yours can schedule meetings for you In the calendar and even in your emails I can also reply on your behalf. So These are tools like Auto GPT or Super AGI is the one that can do this type of work Can. Where the work which an employee does was that type of work you would get an agent to do your Can work as a personal assistant that he will check your mails, reply Will do it. Can auto remind you. These All tasks can be performed by one agent which is a task automation automation agent Is. There are different applications. now like For example be a coding assistant agent What can help developers do? to make you understand that you What do you want to build? In Task Give it a break. help in writing code for that Do it. Helping to debug in testing Of They Are Errors. So for example you are a If you want to make a weather app then for this Front end Back end then API call how to do it how to deploy html css code in text editor Let Small Browser write your output If you help in providing then there are many such We can have all the agents who Increasing our As a Coder Productivity can help for and also different There may be types of agents. As There is e-commerce. Right? In today's date, we are Everybody is looking at us to order something We can order as per our convenience Take it in 15 minutes Delivery in 10 minutes The ziptoto blinks at our house zomato swiggi instam and so many more shopifi you have so many things There are so many e-commerce companies and the biggest There are e-commerce companies which are Generate employment. so here But if we create an agent of our own If you want to do it then for example I do something If you want to buy then here I am telling you should I buy a wrist watch or I want to buy a smart watch. My budget of ₹3000 Is. Reviews that should be good And should have at least 2 day battery Needed So what will this agent do now? Will go to different e-commerce websites and search. will compare everyone's product ratings and I will recommend you the best product He will do it and start telling. so even you get the link can send that this is the website This is the product, here is the link, you can buy it yes. so this is something that the agent is Today we see people in Amazon and Flipkart that we have to buy some stuff. Search There is a box. We will go and search. shows us different things and tells us himself One has to struggle. find ourselves I see how many reviews there are and what they are. But there may be agents who will help you We will easily gather all these information, I will find it, let you know and send you the link. Bus You have to buy it. that is the hole Simplified process that one can do. even a Agent can do the payment as well if they have the Proper credentials of it. Right? So What will every company ultimately do? Agents will develop it because it is much better Will provide user experience. Customer Will give experience. so that is about the E-commerce & Product Research Based Agents. Similar to the health care Be a segment health care assistant agent who can help doctor All the patients who are There is history in summarizing it and giving it Suggest diagnosis based symptoms monitor You can separate health data after that how they don't replace the doctor and idar the nurses but they help like an assistant Right, if there are many patterns then it will help in solving the problem. Extract and analyze data properly There are many types of work which you can do which we can call AI doctor That means the doctor will not replace you But it makes many of the doctor's work smooth. Can make it. can suggest a diagnosis Is. Doctors who prepare patient reports can summarize those reports. I can understand it. so that is all about the health care agent also that everybody The health care company is right now willing to do that so that any doctor who can The platform increases his productivity. Even for the personal travel agent today's date Hospitality is a huge industry in Which is creating so many employments. So here if we want to go to Japan The budget is 80000 for a five day trip right now We have seen many such use cases. By making it perform all the things for you By doing this I am able to give it to you. Now if we If people talk about multi agents then there are many There are such real time use cases. Such as Here we are saying in particular an agent that you have this image of a boy like The position in which it is, in the same way you An image needs to be generated. But the image that you You will generate it, there should be a girl in it who She is reading a book. but in the same pose Like the boy's pose here. Right? Then After that you will see this image here Describe this image with your voice Have to do it. So what is the first step? Post pose determination. So open to it The post model is used here. Then After that it is doing a graft that This is the boy's position. so post to Google to convert image here Or the bit model is being used. than now Convert image to text here Is. So GPT two model is being used. Then The way a girl explains it here Used to be. Look at the way the boy is positioned This was the same girl's position. but now the girl She is reading a book here. Now if you want to do text to speech here then The past speech model is being used. So The different actions that will take place According to the model, it will do your work here And you are getting the final output here. Now because in the AI segment there are billions of Dollars are being invested. every The company wants to become an AI driven company. So there are many such popular AI agents. Which you can try and see. As Baby AGI, Auto GPT, Agent GPT, Crew AI, Manas AI, Lungchen Agents, Super AGI, Meta GPT, Zapier, Zavis by Microsoft has many such agents available which we can go and try using That is how they work. So, that is all about that. but right now the most Which is more important for our career Be it agentic AI, generative AI or such There are segments where there are lots of opportunities Is too much. If we talk about India So India Skills Report or When we talk about employability There is a lot of problem of skill gap That companies need very good talent. Need good data scientists. Good Props They need skill and talent. Good Data Analyst, AI Engineer, Machine Learning Engineers are such a segment which has demand Is too much. but in a proper way There is skill and talent, they are in short supply. So for this It is very important for us to know the new Technology is coming, we will have to study it. Good projects will have to be made. to myself Will have to keep it updated. Then a recruiter will take interest in our profile Yes, I should hire him in particular. so at the same segment of all these things Regarding back to back you will get the skill of All these lectures are from YouTube channel Will keep meeting you. but in addition, if you Preparing for a full fledged job want. where you want every I or Nishant will tell you every concept and topic. Sir, please teach me. You get proper live classes Because this is not possible on YouTube finds it. So for this, the data from The I Scale Premium Science with AI Data Analysis There are also batches in which you can get live doubts Classes and end to end projects will help you In which more than 25 to 30 end to You will get the end projects. Now projects are here for us it is important because your The duration of the interview is very long It lasts for a short while. 15-20 minutes or Half an hour. So in half an hour someone will tell you Also the whole AI, machine learning, deep Learning Python, can't ask. so to you Most questions on projects There are. And only then will anyone hire you When it really works for you. and in fact if If you want to learn the work then you need to You will have to get your hands dirty. You are real Working the same way on time data which will be a data inside the company Analyst, Data Scientist, AI Engineer Or an ML engineer does. So all these works your real time to teach Those projects will be mentioned here Will be. Mock interviews, placement support, Industry recognized certificates, all these Things inside your premium program It is completed in. along with that sometimes Sometimes special discounts are available There are. To check out which you dikkl website or mobile application You can check it out. Related to courses If you need any further advice then click here You have been given the WhatsApp support number Is. You can either call by WhatsApp You can take this. so i'm hoping that that this course about AI agents You must have understood a lot and you will continue to You'll have to try it yourself To keep it updated. so that is all about for Todays video. but going forward now if you Want to know how to create AI agents? If yes then we have more classes for this You can conduct it on YouTube. so as many People also want more focus on AI agents If you find more videos then comment me I go to AI agents or whatever it is Please type it. as many people as The sooner the comments come here the better We will try more to help you and the industry Keep getting videos related to So that your preparation for the interviews For all those things to continue. as an educator After teaching us with such dry throat It is also expected that when we are teaching If you are there then you will really have an impact in someone's life Should be created. Positive Impact should be created and you people I want to crack the placement. Until the interview Have to reach. because that is when we are happiest happens when someone says my placement Done. I got a job. so guys that's all For today's video. I will meet you again In some other video. till then stay curious And keep exploring and keep learning. Thank you.