Welcome to Practical Multi AI Agents and Advanced Use Cases with CrewAI. This is a very practical course, meaning you hear about and learn how to build yourself agentic systems that aren't just shiny technology, but they can be deployed and create real value and application settings right now. Multi AI Agent systems involve multiple AI agents working together to achieve complex tasks by collaborating, delegating, and sharing information. If you took the previous multi-AI agent systems course, you learn that multiple agents can be created, each with specific tasks or roles, and these agents can collaborate to perform complex workflows. In this course, you'll learn how to build agent-based apps for even more advanced use cases, and you see numerous real use cases based on what people are actually doing in practice. When you're building multi-agent applications, a key challenge is balancing the speed and quality of results while also maintaining consistency. Different model choices and sizes can impact these factors. In this course, you'll learn to rigorously test your application by measuring key metrics that you can then use to keep on driving further improvements. You'll also learn to train your agents using human feedback to keep on improving your application over time. It's my pleasure to be here with Joao Moura, who is founder of CrewAI and Joao will be your instructor for this course. Joao also taught the previous course on Multi AI Agent Systems, welcome back. Thank you so much. so good to be back here with you. And with this new practical course on multi-agent systems. We're seeing so many companies building multi-agent AI applications, running in aggregate, tens of millions of agents using CrewAI. And creating more complex agent workflows that involve multiple agents working parallel, multiple crews as well, and doing so much rigorous performance testing and training steps. So, Joao, in the previous course you had explained and demonstrated the basic building blocks of multi-agent systems, things like how agents can work together on the tasks, how they can use tools, how you can build some really cool applications with relatively few lines of code. Maybe you can say a bit about what learners can look forward to in this course. Sure, we will start with a quick review of agentic systems. You will then learn how to integrate your multi-agent automation with internal and external systems. This enables your app to perform actions like query internal data, calling existing systems, sending emails, and so much more. Next, you'll learn how to create a variety of crew setups from sequential to parallel and anything in between, including a few hybrid setups. You will learn how to run tasks in parallel, and also how to connect multiple crews in the pipeline. Then we will look at how you can optimize the performance of your multi-agent systems using testing and training methods. After that, you will learn how to create crews that employ multiple different LLM models to complete a task. For example, you could have a researcher agent using a smaller and faster model for a relatively simple task, and the writer agent using a large model that has been fine tuned to reflect your company's brand voice. I will be referring to this as a multi-model approach that lets you mix and match multiple LLM models based on the task at hand, which can help you build a more efficient and customized AI system. Many people have worked to create this course. I'd like to thank the whole CrewAI team and from DeepLearning.AI, Esmaeil Gargari and Geoff Ladwig have also contributed to this course. This course is going to be a lot of fun. You will build several practical apps like an automated project planning, lead scoring and engagement automation support, data analysis and content creation at scale. I've really enjoyed using CrewAI myself and I'm confident that you will too. Let's go on to the next video to get started.