ABSTRACT
Developing efficient and reliable software is a difficult task. Rapidly growing and dynamically changing data sets further increase complexity by making it more challenging to achieve efficiency and performance. I present practical and powerful abstractions for taming software complexity in this domain. Together with the algorithmic models and programming-languages that embody them, these abstractions enable designing and developing efficient and reliable software by using high-level reasoning principles and programming techniques. As evidence for their effectiveness, I consider a broad range of benchmarks involving including sophisticated algorithms in geometry, machine-learning, and large-scale cloud computing. On the theoretical side, I show asymptotically significant improvements in efficiency and present solutions to several open problems using the proposed techniques. On the practical side, I present programming languages, compilers, and related software systems that deliver significant improvements in performance, usually with little effort from the programmer.

Conference information