Do you want to know what my first experience with generative AI was? Wildly inconsistent results. Half the time, the response wasn’t in JSON so I could parse it, and when it did work, it was slow. The problem? My prompts were bad. So, I started figuring out how to make them better.

What structure works best? How to determine the best temperature level? Should I care about function calling? Will my prompts perform well with 100 users?

The more I figured out how to use generative AI, the more I realized I was becoming a “prompt engineer”. You know, the thing that’s supposed to replace all developers in the future. I’m not sure about that. But in the AI era, we need to know how to work with it.

If you’d like to skip all the trial-and-error steps, join this session to learn how to take your prompts from a simple request that returns wildly inconsistent results to a complex prompt that makes your new AI features more stable and reliable.