Steven Gustafson
 Chief Scientist, R&DataScience,Maana

Steven Gustafson is Chief Scientist overseeing research and data science at Maana. Previously, he spent over 10 years at General Electrics Global Research Center in New York, developing novel solutions for massive-scale, click-stream log mining for media and industrial equipment businesses, open source intelligence, and network analysis for the intelligence communities; and knowledge-driven data and analytical solutions across the finance, healthcare, and industrial sectors. Steven drove several company-wide strategies for Big Data, Semantics, and most recently Artificial Intelligence. He founded the Knowledge Discovery Lab at GE and co-founded the journal of Memetic Computing, where he was the Technical Editor-in-Chief.

Steven has chaired conferences and program committees, and serves on several editorial boards. Steven is a former winner of the IEEE Intelligent Systems AIs 10 to Watch award, holds over 10 patents, and has authored more than 40 peer reviewed articles.

Steven earned a PhD in Computer Science from the University of Nottingham, UK, and was a research fellow in the Automated Scheduling, Optimization and Planning Research Group. He received his BS and MS in Computer Science from Kansas State University, and served as research assistant in the Knowledge Discovery in Databases Laboratory.

Optimizing Oil Shipping with a Computational Knowledge Layer

Enterprises are focused on enabling digital transformation and optimization of their operations and business decisions. The strategy for digitization involves the capture and reuse of knowledge that is used to make decisions and refined through experience and learning. Knowledge graph technology is helping to address digitization by providing critical context and interoperability of data. The optimization of decisions, which usually includes an analytic or machine learning model, can result in a new form of silo within the enterprise. In this talk, using a real-world, customer case study of Oil Shipping, I describe how we can break down the analytic model silo with the creation of services and functions that implement key decision logic over the knowledge graph to form an enterprises computational knowledge layer.