video1.0<iframe src="https://www.loom.com/embed/f9814fb1708c43b58f5c9937cd85dd6c" frameborder="0" width="1664" height="1248" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12481664Loomhttps://www.loom.com12481664https://cdn.loom.com/sessions/thumbnails/f9814fb1708c43b58f5c9937cd85dd6c-a80d58f62ad020c6.gif173.71In this video, I reflect on the mistakes I made while building AI applications for ArcGIS, Gemini, and Google Cloud. I discuss the importance of rich logging for prompts and model interactions, the balance between cost management and delivering value first, and the significance of defining and tracking success metrics. I emphasize the need to focus on building the product first, then optimizing for cost and efficiency. No specific action is requested from the viewers.