Dr. Nohadani discusses the latest discoveries and future direction of advanced analytics, optimization, and machine learning.
BOSTON (PRWEB) MARCH 19, 2020
Dr. Nohadani is an optimization and machine learning scientist with a proven track record of developing algorithms and solutions for big-data environments affected by uncertainty and subjected to complex risk measures. He sits down to talk about how the latest innovations in data science provide plan sponsors with powerful new tools that help improve the visibility and analysis of their data and prescribe actions to improve their health plan.
1. What was your role prior to coming to Benefits Science Technologies?
A: Most recently, I worked as an Associate Professor of Industrial Engineering and Management Sciences at Northwestern University. In my research group, we studied dynamic optimization and analytics under uncertainty and risk, with applications in healthcare, supply chain management, analytics, and technology. In particular, we focused on data-driven dynamic systems that rely on human interaction or have decision-makers with complex risk requirements. I received my MS degree in physics from the University of Bonn, Germany and my Ph.D. in physics from the University of Southern California, Los Angeles. I then joined the Operations Research Center at MIT as a postdoctoral researcher, developing novel robust optimization algorithms. As a research fellow and instructor at the Department of Radiation Oncology at Harvard Medical School’s Massachusetts General Hospital, I conducted mathematical research on cancer treatment.
2. What is your role at BST?
A: I lead our data science team of highly talented Ph.Ds. and MSs., who work on core components of our products. Our research team integrates the cutting-edge developments from the scientific communities of optimization and machine learning with the requirements of practitioners to create methods and ultimately products that empower data-driven decisions beyond individual expertise. We leverage the newest research findings to not only analyze the plethora of data sources at BST but also provide predictive and – more importantly – prescriptive insights. This allows for optimal resource allocation and targeted treatments in a timely fashion while maintaining costs.
3. You’ve had a pretty varied career path, what attracted you to BST?
A: My goal as a data scientist is to push the boundaries of science and technology in order to expand our knowledge and to create solutions that affect the world positively. This endeavor typically requires three components: exposure to relevant problems, proximity to cutting-edge science, and access to the real-world, e.g., through observation and data. I have pursued this goal in academic settings for a number of years and realized that while the first two components are available, the access to real-world data is limited. Alternatively, many private organizations pose relevant questions and have access to observations but are often not closely connected to the newest scientific discoveries. At Benefits Science, I was delighted to find a productive combination of all three components in a unique setting as a research-driven endeavor with a mature business model, providing the ideal format to innovate for real-world settings. Moreover, the societal relevance of healthcare elevates the motivation on a personal level and its economic impact promises transformation.
4. Where do you see these capabilities being applied in the healthcare decision space?
A: Benefit plan design is traditionally centered on the expertise of decision-makers. Given the complexity of these products and possible combinations, plan designers resort to last year’s choices to inform the next designs in an iterative fashion. At BST, human expertise is built into an algorithm that leverages all data to provide the optimal design that meets all desired criteria while minimizing costs. BST’s technology is able to efficiently solve large-scale mathematical robust optimization problems that were not possible before. Similarly, risk management is currently guided by broad economic trend analysis and is rarely tailored to the studied cohort. At BST, cohort related factors are also accounted for to not only predict future risk factors of individual members but also to prescribe which proactive steps could advance members’ health and mitigate risks. This could only be made possible at BST by integrating cutting-edge machine learning techniques with the newest optimization algorithms.
5. Interesting, can you give a specific example of how that works?
A: Employers often offer additional wellness and preventive benefits. However, budgetary considerations constrain the portfolio of these benefits. The prescriptive nature of BST’s technologies allows us to optimally tailor the portfolio to a subgroup of employees who have different needs, instead of the current one-size-fits-all approach. For example, members with musculoskeletal problems will be distinguished from those who require dietary support and our methods would help employers best address those needs while maintaining budgets. Moreover, we leverage past performance data to recommend the set of providers who would optimally improve health outcomes.
6. How do you see these types of advancements impacting the future of healthcare analysis?
A: All aspects of healthcare-related decision making, from medicine, delivery, policy, supply chains, and economics are becoming increasingly more data-driven. At BST, we are developing solutions that affect many of these directions. Since accuracy, scalability, and interpretability is at the core of our research and products, I believe we are well equipped to address future challenges. For example, our interpretable care pathways are trained with past data and informed by experts’ knowledge, hence well positioned to curb the increasing occurrence of diabetes in a proactive fashion, improving patient’s health, reshaping treatment schedules, and managing costs. The abundance of data and the emergence of powerful methodologies makes it an exciting time to be a data scientist in healthcare.
About Benefits Science Technologies
BST provides data analytics software/analysis to help manage the risk of self-funded health plans. Plan holders improve their connection to data, empowering optimal decisions to control costs, and improve the quality of care for plan members. The company distributes these capabilities to employers directly, through brokers/consultants, or by association with carriers (stop-loss and voluntary benefits providers), benefits administrators, professional employer organizations, third party administrators, pharmacy benefit managers, and/or cost-containment providers.
Founded in 2012, the company is recognized as a world-leading research and applied science team, developing advanced analytics and robust optimization methodology for complex health insurance decisions. Our mission is to improve health outcomes while lowering costs by helping our clients uncover the details of their health plan – and providing actionable insights along with outcome predictions for the recommended decisions. Our strategy is to build world-class solutions based on cutting-edge scientific insights.
Benefits Science Technologies
Mark Hufham [email protected]