DAVID DEVILBISS, PH.D.
Director and Data Scientist
**** ***** ******, *******, **, 53705 Mobile: 608-***-**** **********@*****.***
Accomplished Director and Medical Data Scientist with twenty years’ experience innovating, implementing, and commercializing end-to-end platforms for acquisition, storage, distributed analytics, and visualization with a deep knowledge of high dimensional-multivariate statistical models, machine learning, and predictive analytics.
Core Competencies
Enterprise-Level Project Management
Global Strategic Alliances
Emerging Technology Implementation
Product Lifecycle/ Management
IT Infrastructure Design
Data Standards and Curation
Forward and Inverse Statistical Models
Machine Learning and Data Mining
Predictive Analytics
Parallel computing
Experience
Founder and Principal Data Scientist
NexStep Biomarkers
June 2007 - Ongoing Madison, WI
Built all signal processing, data mining, and machine learning algorithms, statistical analyses, and software platform that has processed >5TB of brain-related time-series data.
Directed global business development and operations of all data projects, strategic partner engagement, customer contract negotiations, and quality assurance/quality control increasing margins 10 fold.
Technologies
Matlab, R, JMP, MySQL, Parallel computing, Clustering (K-means, Gaussian mixtures, hierarchical), Dimension reduction (Principal/Independent component analysis, Factor analysis), Multiclass prediction, Fault tolerant mathematical modeling.
Principal Investigator and Researcher
University of Wisconsin
June 2007 - Ongoing Madison, WI
Developed novel signal processing algorithms, mathematical models, and big data processing strategies to analyze brain-related time-series data for publication in 11 peer-reviewed journals.
Directed Scientific Research on federal projects (NSF/NIH) totaling $3 million.
Awarded Critical Compensation to retain technical expertise.
Lead and mentored 3 technicians and graduate students in computational neuroscience.
Educated an average of 20 graduate students and post-docs/year in cutting edge machine-learning approaches through HAMLET (http://psych.wisc.edu/devilbiss/Hamlet/)
Technologies
Matlab, R, Excel, Generalized linear models (Boosted, L1, L2, and Elastic net), Clustering (K-means, Gaussian mixtures, hierarchical), Dimension reduction (Principal/Independent component analysis, Factor analysis), Multiclass prediction, Machine Learning, Fault tolerant mathematical modeling, Data visualization.
Founder and Head of Research and Development
Cerora Inc.
February 2012 – November 2016 Bethlehem, PA
Built signal processing, mathematical models, and predictive analytics to aid in the diagnosis of Alzheimer’s disease and mild traumatic brain injury resulting in 25 scientific abstracts and papers and multiple patents.
Created and maintained advanced clinical EEG/cognitive testing client and cloud platform providing brain diagnosis information to doctors within 3 minutes.
Led cross-functional bioinformatics team to commercialize scalable clinical diagnostic web application platform resulting in the company’s valuation of 12 million dollars.
Technologies
Matlab, R, Python, JMP, Predictive Analytics, Machine Learning, Fault tolerant mathematical modeling, Data visualization, Database Design and implementation, Multivariate Statistical Analyses, C/C#, PHP, JavaScript, HTML, SQL and NoSQL Database Design and implementation.
Education, Publications, and Patents
Postdoctoral fellow in Computational Neuroscience (2006) University of Wisconsin, Madison, WI
Ph.D. in Neuroscience (2002) Drexel University, Philadelphia, PA.
B.A. in Psychology (1994) University of Delaware, Newark, DE
Current Publication and Patent List: https://Goo.gl/ihje1s
Additional Expertise
Multiple Programming Languages (C/C++/C#, PHP, JS, HTML, XML)
Linux SysAdmin and Bash Scripting
Database design
Apache, NodeJS, MapR, AWS, S3, REST, SOAP
Advanced Statistics (multivariate linear and non-linear regression, decision trees, exploratory data analysis, modeling, and neural networks)
Data Visualization (Matlab, ggplot, SpotFire, Chartjs)