Daniel Crankshaw

TL;DR: This paper introduces GraphX, an embedded graph processing framework built on top of Apache Spark, a widely used distributed dataflow system and demonstrates that GraphX achieves an order of magnitude performance gain over the base dataflow framework and matches the performance of specialized graph processing systems while enabling a wider range of computation.

TL;DR: Clipper is introduced, a general-purpose low-latency prediction serving system that introduces a modular architecture to simplify model deployment across frameworks and applications and improves prediction throughput, accuracy, and robustness without modifying the underlying machine learning frameworks.

TL;DR: Clipper as mentioned in this paper is a general-purpose low-latency prediction serving system, which is able to meet the latency, accuracy, and throughput demands of online serving applications by introducing caching, batching, and adaptive model selection techniques.

TL;DR: The challenges and architectural considerations required to achieve this functionality, including the abilities to span online and offline systems, to adaptively adjust model materialization strategies, and to exploit inherent statistical properties such as model error tolerance, all while operating at "Big Data" scale are described.

TL;DR: It is demonstrated that GraphX achieves comparable performance as specialized graph computation systems, while outperforming them in end-to-end graph pipelines, achieving a balance between expressiveness, performance, and ease of use.