One Tool — 100+ AI Agents (tutorial for beginners)

AI agents are absolutely taking over the industry. And trust me, I'm not joking. They can do everything now from helping your business grow to simply streamline your workflows. And the best part, you can create one just a few minutes for next to nothing. No more juggling multiple subscriptions. One tool is all you need for pretty much any AI task you can think of. In the last video, I already showed you most of what you can do with Chat LM by Abacus. There's one thing I promised to show you more of AI agents. Now, normally if you want to create an AI agent, you have to learn how to use APIs, find a platform that lets you connect different tools and then spend hours debugging everything. It's a lot of work. But with Chat LM, there's one feature that makes it super easy, AI engineer, and you don't need to deal with any of the usual stuff. You create an AI agent directly through a chat. No coding required. And it all happens super quickly. You ask for an agent and boom, you've got one. Here's how it works. First, I open the uh AI engineer tool. Then I have two options. Create a custom chatbot or create an AI agent. With these custom chat bots, you can set them up for your website or your team. Train them on your own data and connect them to your own data set. This way, the model always has access to the most relevant up-to-date info. It's a gamecher for things like uh supercharging your sales team, helping HR handle more tasks, or even speeding up customer service. Businesses are already using this in a big way. And if you visited any website in the past six months, you've probably encountered a chatbot that works just like this. We are going to focus on the second option, creating agents. It's actually super easy to do this. All I need to do is select the template from the list of suggested ones. Then I just enter some custom details about the agent I'm creating, like its name and any specific instructions. The custom instructions are easy. I simply write what I want the agent to do. And trust me, I really recommend doing some planning beforehand. Grab a pen and paper and sketch out basic flow of your agents behavior. Think about how you want your agent to act, what actions you should take one by one and how they all connect. It doesn't have to be detailed or a complicated scheme. Just focus on the basic logic behind it. Once that's done, you can describe it in the input field. For example, you might write an agent that takes a list of sports teams as input, then searches the web for upcoming games. It should output the times of the games and details and how to watch them. After I submit all this info, AI engineer asks me about the steps the agent will perform. It's already figured out the basic steps and I have the option to add, remove or adjust them, including changing the description of each step. Here you can create pretty complex workflows as long as you can break them down step by step. Then I just wait. Chat LM handles all the coding, all the logic and the reasoning needed to figure out how the agent should work internally. Once it's done, gives me two links. one to try out the agent right away and another to configure its elements. The second link opens up an editing window like the ones you find in most AI agent creation tools. But the cool part is you don't even need to go there. Just click the first link and your agent is ready to go. That's honestly what I usually do. And don't even say it was a difficult process. The whole thing took me like less than 2 minutes from the idea to the fully functional agent. Sure, I'll need to spend some time connecting the necessary data sets and knowledge sources, but that's a breeze compared to all the work the AI has already done for me. And the agents I created with Chat LM aren't just simple chat bots. They come with something called artifacts. Artifacts are like many apps within the AI itself. For example, an artifact can access another AI like Opus for brainstorming ideas and writing stories or Smog for managing data and creating interactive dashboards. And the best part, I can create my own artifacts. Super simple and easy. Just like creating agents, I do it through the chat. For example, I could just ask design a weather dashboard that shows the current temperature and forecast. AI takes care of all the coding and then I get an interactive view of the artifact that I can tweak and customize directly in the chat. I can also create things like a to-do list app where I can check off tasks where it could be a quiz app, a puzzle game, or even a portfolio website. all within the same interface and using the same models you're already familiar with. Let me show you a few examples of the kinds of agents with artifacts that you can create in Chat LM. Here's a patient engagement planner. First, I can select the clinic services I'm offering, then choose the goals I want to achieve and specify the patients demographic and literacy level. All the input fields are customizable. Some are large text boxes, others are check marks with multiple choices, and some are single choice options. And all of them work together seamlessly to collect the right data. Here's another cool one. Customer support agent that takes in a customer's message, attachments, and relevant HR info, and then creates a formatted email for the HR rep to simply copy and paste. It's a really basic agent, but it demonstrates how much flexibility you have, or in my case, how much freedom I have to design exactly

what I need. Now, check this out. This one's a few levels above everything I've shown you so far. It's a customer account manager that takes customer data and sends it directly to Salesforce, which is an entirely separate platform. And the best part is the agent works perfectly placing the data exactly where it needs to go without any mistakes. So really, if you're careful and you have a bit of time to spend, you can create some pretty complex agents that will seriously speed up any workflow you have. I'm actually thinking of creating an agent that takes the 50 most popular YouTube videos on a specific topic and gives me a list of ideas with thumbnails, titles, keywords, and even full scripts. But actually, I don't even have to do that because there's already a powerful agent right inside Chat LM. It's called Deep Agent Pro and it's really powerful. It can build apps, write reports, create presentations, and automatically connect to all your systems and perform agentic tasks. With Deep Agent, I can do detailed data analysis and create interactive dashboards in a click with Deep Agent figuring all steps out on its own. Creates apps and I can even set custom domains for hosting these apps. I'll admit the prompt used is quite long and you do need to know how to prompt to get that answer. Then I answer a few questions from the agent, give it a minute to think, and voila, here's a fully ready app. Deep agent did all the coding, debugging, all the tests fully autonomously. I didn't even lift a finger. And games. Deep agent can also create games like the game of life. It's a simple game, but Deep Agent easily created a website that works great in the game that's interactive, has solid logic, and overall looks great. Deep agent can also do tasks that require browsing like finding cheap plane tickets. That is a serious next level agentic workflow. And it can seamlessly connect to Google Workspace and work with art data for creating apps, websites, doing data analysis, etc. And it works with Slack too. Slack and Gmail. So I could automate my work pretty well if I wanted to. As I mentioned before, when it comes to agents, there are a lot of LMS out there, each with its own strengths. Chat LM isn't just a platform for building agents. It's also a one-stop hub for the most popular AI models that you can access with just one subscription. You've got Chad GBT4, 4. 1, Gemini 2. 5, Claude Sonnet 3. 7, Grog 3, Lama, and much more. All included in one subscription and all available under one platform. What's even better is that when you're using these models, you don't have to worry about manually switching between them to get the best results. Chat LM comes with a built-in feature called root LM, which makes all the decisions for you. It analyzes your prompts and automatically selects the model that will perform the task the best. For example, it might stick with Chad GBT for general writing tasks, but switch to a model like Gemini for analyzing specific data. All the models here are fully packed with features. There is still multimodal, meaning they can analyze documents, work with images, generate new files and reports, dive into deep research on papers, and much more. For instance, you could upload a PDF and your agent will read it, answer questions about the contents, summarize it, and even generate interactive graphs based on that document. Now, all this already sounds like a pretty solid deal. But if you're running a business, you might be wondering how well Chat LM integrates with your existing workflow and internal systems. Let me give you a quick list of things you can connect with Chat LM. Select Teams, Google Drive, One Drive, and that's just the beginning. You can create custom agents to handle project management, automate internal communications, organize your documents, and more. It all seamlessly integrates with your existing systems requiring minimal effort on your part. And even if you decide not to create any agents, there are still plenty of useful features you can take advantage of. For example, if you're a developer, there's a feature called Code LM that you can add to Visual Studio. Works like a plug-in helping you write and debug code in real time without having to switch apps. You can also connect it directly to your GitHub repository, so it always has the most context and give you the most accurate answers. Sure, it won't replace a human developer, but it will definitely make your work a lot faster and more efficient. And all this is included in just one subscription that's actually cheaper than any individual model or platform I've mentioned before. For example, a Cad GBT subscription will cost you $20 and Gemini will be the same at another 20 for Claude, 30 for Grock, and who knows how much for all the different agent building platforms this one replaces. When you add it all up, you're easily looking at hundreds of dollars saved every month. And I can even send referral links to my friends which makes their subscription cheaper and in turn adds a little bonus to my account. So I can pay for my subscription without spending a single dollar of my own money. I have the option to pay monthly or yearly. And if for any reason I want to close my account, I can do it without any hassle at any time. There are no features locked behind pay walls. Just one subscription and I can do everything. I will leave a link in the description. So make sure you check it out. Thanks for

watching and I'll see you in the next video.