AI Solutions Design Mini Course

Hey everyone, and welcome to this incredible 10-step microcourse. This is an hour-long deep dive whiteboarding presentation that I did for an event. People paid $425 to attend, and they were happy to do it. We had 98% customer satisfaction, um, which means someone must have graded me, I guess, a 90. I don't know, uh, but over 100 people attended, so This is your chance to kind of be in the room where it happens. And I wanted to quickly talk you through what is in this course. So you can see here in this course, I take you through the classic AI design architecture session that I give to executives. And as I mentioned, I taught some solution architects how to do this at an event recently. And so I recorded it and cut it up just for you all. In this course, you can see the first— we're going to talk about How to really have a conversation with a leadership team or an executive team. How to then craft an AI solution. How to go from a problem identification and metric determination to pick the right problem to solve with the right AI solution. We then talk about data. We then go through model evaluation and selection. I talk about the different types of models. We also go through how to make sure you're using a model that is not predisposed to bad behavior. We also talk about latency and cost and the trade-offs that come with that model selection. And then we end talking about the most important aspect, which is how do you make sure your models don't— or behave well over time? How do you make sure that they scale? And that scalability comes down to 3 things: security, accuracy, and trust. And so in this session, I dive deeper into what those 3 elements look like and how you can ensure that you have a project that ultimately will last the test of time. So hopefully you enjoy this little mini crash course and getting a seat inside a secret room where I divulged all of the things that I use, learn, and teach about how to design responsible AI at scale. Enjoy.