Okay, so we just built an amazing crew. By now, you're probably getting excited and I am as well. But we can take a step further. In this lesson, we're going to build a multi-agent system that's going to be able to help us plan an event. Think about a meetup or a conference, whatever it might be. These agents are going to help us to find a venue, do the planning, planning around the catering, and everything that comes with that. So let's jump into it and see what that looks like. Welcome to our next lesson. In this lesson, we're going to be talking about tasks. I'm super excited about this one. We talked about so many things already, and this is another great topic for us to dig into, because at the end of the day, if you think about your agents, they are performing tasks. So tasks are a cornerstone on multiple agent systems as well. So when you talk about tasks, I want to bring back something that we talked about a few lessons ago, the manager analogy. You might remember that we discussed that this manager mental framework to create agents, where you put yourself into a manager's shoes and you think about what would be the people that you would hire in order to do a series of tasks or the job that you're trying to accomplish. And then you kind of use that as a starting point to define what will be the roles, the goals, and the backstory of your agents. So you need to have a very good understanding of what is the goal that you're trying to accomplish, and you also need to have a very clear understanding of what is the process throughout to achieve this goal. So with that in mind, you're gonna have a clear understanding of the people that it would hire, and those will be your agents. But now, I want to build on top of this analogy. I want to expand this mental model to how you think about your tasks. Because whenever you hire someone to do something, you also cease to think about how you're going to delegate work to them. Especially if it's more of a junior person, you want to make sure that you're going to delegate work to them. Especially if it's more of a junior person, you want to make sure that you're very explicit about what you expect that they do and what is the expected result. So I would say for you to build great agents, you should add tasks I expect individuals on my team to do. And you can use that to create the tasks. And whenever you're creating a task, again, you need to have at least two things top of mind. One, a clear description of what is the task. And two, set a clear and concise expectation. Again, to go back to our analogy, let's think that you just hired a junior engineer to work on your team and you're tasked with coaching this person. It's their first day on the job. You want to give them a specific task to work on. You want to explain that task very truthfully. And then you want to make sure that you say what you expect them to do. to make sure that you say what you expect them to do. So, query AI kind of forces you to think that way because query AI makes sure that it's mandatory for you to set at least two attributes on every task that you create. The description and the expected outcome. And that is used to build better internal prompts. But that also applies to any other framework out there. You will get better results if you go the extra mile on explaining what is a task and what you expect your agent to do. Other than that, there's a bunch of other things that you can set on the task. Query AI offers a bunch of hyper-parameters that you could use, like setting a context, for a bunch of hyper-parameters that you could use, like setting a context or setting a callback or overriding the agent tools with specific task tools, force human input, kind of like making sure that before the agent finishes the work, it stops and asks you how you feel about it so you can give further instructions. Also, we can talk about execution these tasks asynchronously or outputting their results as PyDentech objects or JSON objects or even outputting them as a file or running these tasks in parallel. So you can see how there is a bunch of options when we're talking about tasks in here. And you're going to need to take all that into account depending on how complex your multi-agent system is. Crew offer all these options and you're going to find a bunch of them in other frameworks as well. Maybe not all of them and maybe in a different shapes and forms, but in the end of the day it's all about how you can set up your agents for success to perform their tasks super efficiently. So with that in mind, let's jump into code real quick.