Deep learning algorithms are now transiting from proof-of-concepts or academic research to production deployments in industry. In this talk, we discuss lessons learned in engineering algorithms at scale, from model development to inference optimizations to performance monitoring in the field. We also present the Intel Nervana portfolio of software and hardware to enable faster deployments.
Hanlin Tang is a Senior Algorithms Engineer at Intel, and previously from the deep learning startup Nervana Systems. At Nervana, he worked on developing the open-source framework neon, building deep learning models for computer vision, and contributing to the Nervana Graph project. He received his PhD from Harvard University, where he investigated the role of recurrent neural networks in human cortex. His main interests are at the intersection of neuroscience and deep learning. Hanlin previously was at RAND Corporation, where he built quantitative models to study national security issues.