Virtually every human endeavor is encountering the need or opportunity to obtain new insights through the analysis of massive datasets. For maximum efficiency, it is necessary to employ both parallelism and fast algorithms, as is done in physics for N-body simulations. Each of these aspects (parallelism and fast algorithms) greatly complicates the development of data analysis and mining code, not to mention the inherent complexity of many statistical and machine learning methods. This is evidenced by the very small number of codes in existence which implement statistical/machine learning methods using fast algorithms on parallel machines. In this talk, Aparna Chandramowlishwaran will present Portal, a new domain-specific language and compiler designed to enable high-performance rapid implementations of data analytics and mining tasks with minimal effort. The goal in the development of Portal is three-fold, (a) to enable scalable, fast algorithms which have O(n log n) and O(n) complexity, (b) design an intuitive language and API to enable rapid implementations of a variety of problems, and (c) to enable parallel large-scale problems to run on current and future architectures. The initial target includes problems such as k-nearest neighbors, kernel density estimation, expectation maximization, to name a few which are commonly used in various analytics tasks. The language and intermediate algorithm representation are independent of the architecture, making the approach portable and easily extensible for different platforms.