As scientific data is reaching exascale, scalable and energy efficient data analytics is quickly becoming a top notch priority. Yet, a sustainable solution to this problem is hampered by a number of technical challenges that get exacerbated with the emerging hardware and software technology trends. In this paper, we present a number of recently created "secret sauces" that promise to address some of these challenges. We discuss transformative approaches to efficient data reduction, analytics-driven query processing, scalable analytical kernels, approximate analytics, among others. We propose a number of future directions that could be pursued on the path to sustainable data analytics at scale.