So let's do a quick recap on everything that we learned so far. So we learned that this mental framework for creating agents is super helpful because it helps us to understand what are the agents that we're going to create and also what are the tasks that we expect them to do. We also learned that when we're thinking about these tasks, we should be very much thinking about what are the clear descriptions and what is the clear expected output that we want to get for each and every one of them. And we also got the chance to mess around with a lot of those individual attributes that the tasks offer, from parallel execution to output JSON to use Pydantec, and a lot of different interesting stuff. You're going to get the chance to play around with all those different arguments as you build the systems yourself in the future, but I hope this was a good way to showcase how powerful these systems can be. So I really appreciate you sticking by in this lesson and going through this. I think this is a turning point where you kind of start to build really production case agents where you understand how they work and it's only going to get better from here forward. So congratulations for you for sticking around and being learning about this. I would say pause the video a little bit, go back into the Jupyter notebook, change it a little bit, see what results you can get from it. It's going to be a lot of fun and I see you in the next lesson.