Dr. Vikas Agrawal
Sr. Principal Data Scientist
Member since 6 months
Dr. Vikas Agrawal
Specialises In (based on submitted proposals)
I currently play the role of a Senior Principal Data Scientist at Oracle Corporation in Cognitive Computing for Oracle Analytics Cloud. My current interests are in automated discovery, adaptive anomalous trend detection in non-stationary streaming multi-variate datasets, intelligent context-aware systems, and explaining black-box model predictions, building on previous work that created:
- Activity context-aware virtual personal assistants with N=1 personalization for insurance, pharma, retail and investment banks with risk and fraud detection
- Real-time asset management, predictive maintenance, and reliability risk prediction systems using Internet of Things (IoT) driven data-streams in mining and production systems for eliminating down-time, waste, and delay in manufacturing with explanability.
- Automated discovery, anomaly detection and guided modeling systems for scientific analysis of HCM, CRM, ERP and SCM datasets
Before Oracle, I worked as Principal Data Scientist for the Cloud and Big Data business at Infosys, and as a Principal Research Analyst at the Center for Knowledge-Driven Intelligent Systems at Infosys’ Enterprise Technology Research Labs. Prior to joining Infosys in mid-2010, I coordinated multi-disciplinary research and development efforts as a Staff Engineer (Silicon Integration, Design-In Quality, and Reliability) within Intel Corporation‘s Technology Development (Folsom, CA and Chandler, CA) organization for five years. I was fortunate to work on post-doctoral research at California Institute of Technology (Pasadena, CA), anchoring development of mathematical models jointly with researchers from Jet Propulsion Labs and UC Irvine for the Computable Plant project funded by NSF FIBR.
I received a B.Tech. in Electrical Engineering from Indian Institute of Technology, New Delhi (1997), an M.S. in Computer and Information Sciences and Ph.D. in Computational Mathematical Modeling / Statistical Modeling in Systems Biology from University of Delaware, Newark, DE. I had the privilege of conducting research with AstraZeneca‘s International Bioinformatics Target Discovery team for identifying targets for diseases of the central nervous system and with DuPont Pharma’s Stine Haskell Research Center (now part of Bristol Myers Squibb Company) to create datasets for FDA approval of the anti-HIV drug SUSTIVA (Efavirenz) for short periods of time.
Over time I have been working on developing solutions to three grand challenges:
a. Zero Down Time, Zero Intrusion, Zero Loss: Can we provide highly reliable temporally relevant information in people’s work context on their mobile devices today? What if we can prevent equipment failure by predicting maintenance or replacement needs? What if we can give pre-summarized predictions and recommendation with seamless data provenance to technicians? What if we can predict risk for a fire by connecting the condition of a boiler, simulating it with the ambient?
b. Zero Waste, Zero-Delay – Just In Time Supply Chain: Can our sensors in the supply chain prevent food waste, Rx/Dx misuse, detect counterfeiting, predict manufacturing requirements from orders online, and propagate the requirements up to the supply chain all the way to suppliers?
c. N=1 Personalization: What if we present information and enable actions relevant to the context (location, role, social circumstance, access level) of the users? Can we provide adaptive home security to learn from regular family behaviors, and detect the unusual through heat, light and audio signals? Can we minimize energy loss and optimize energy usage by predicting when certain water or air temperature patterns are needed? What if people can seamlessly use our devices to build consumer services accessible from ubiquitous mobile devices?