A disproportionate cost of weather-related hazards is related to "extreme" weather events, which occur infrequently enough that even in the current climate we have insufficient observations to fully characterize just what kind of weather is fully possible. The possible change in the frequency, intensity and duration of these already poorly constrained events due to climate change means that there is also a disproportionate interest in the changing costs associated with this kinds of events. At Jupiter Intelligence, we build products to answer questions about changing weather extremes. We employ emerging methods and datasets in the climate community that balance customer requests, the state of the science, and global scalability. In this talk, we present an overview of the kinds of questions customers are asking about extreme events, as well as some methods we've chosen to use at Jupiter to address them. Given the wealth of extreme value analysis methods available that have been considered in weather and climate applications, we outline pathways for mapping user requests to what is scientifically defensible and globally consistent. A particular focus is given to the user interpretation (or misinterpretation) of extreme value analyses, and how our methods have been shaped to respond to this important consideration. Finally, ongoing challenges where we are interested in seeing more focus in the weather and climate extremes community will be presented.