I am a researcher and venture technologist who has developed numerical algorithms for multi-national corporations, research universities, and most recently start-ups. I have developed solutions for finance, mobile applications, and earth sciences and am passionate about solving data-related problems in industries that least expect it. I am founder and principal of PolytopX, a consultancy focusing on novel
I am a researcher and venture technologist who has developed numerical algorithms for multi-national corporations, research universities, and most recently start-ups. I have developed solutions for finance, mobile applications, and earth sciences and am passionate about solving data-related problems in industries that least expect it. I am founder and principal of PolytopX, a consultancy focusing on novel approaches to cloud computing applications for learning and uncertainty quantification. I am also co-founder at glowfi.sh that delivers a generalized machine learning service for connected devices to reduce the barriers to entry for companies lacking data science expertise.
CEO & Co-Founder @ Empowering IoT with an integrated API-machine learning solution that doesn't just make data intelligence easy, it makes device intelligence easy too From November 2014 to Present (1 year 2 months) Principal and Founder @ I founded PolytopX for two reasons: to spend more time thinking about new approaches to classic optimization, uncertainty quantification, and analytics problems and to be able to work with a variety of companies helping them solve these kinds of R&D problems.
Current projects and interests include financial market data analysis and algorithmic trading, uncertainty quantification (currently exploring uncertainty in estimated I.V. surfaces and their effect on options pricing), geophysical imaging problems (high-dimensional stochastic imaging utilizing sub-spaces and non-linear learning models) history-matching (Stanford Center for Reservoir Forecasting collaboration with T. Mukerji), and a variety of learning problems. From 2012 to Present (3 years) san francisco bay areaCo-Founder @ Wine4.Me is a smartphone app created by VineSleuth which uses sensory science, personal preference algorithms and predictive analytics to empower casual wine drinkers and help them find more wines they are sure to enjoy. From April 2011 to March 2015 (4 years) San Francisco Bay AreaPrincipal Researcher @ Development of geometric sampling techniques for efficient uncertainty quantification in nonlinear optimization problems.
Improvement of polynomial chaos efficiency using compact quadrature From 2007 to 2012 (5 years) Senior Geophysicist @ Implementation of several interpretation techniques for marine electromagnetic data including use of dense-measurement array acquisitions for coherent noise reduction and signal enhancement
Designed 10+ marine CSEM field acquisition surveys and led data interpretation. From 2005 to 2007 (2 years) Senior Research Geophysicist @ Development of multi-dimensional imaging and modeling algorithms
Investigation of novel data acquisition geometries for determining parameter anisotropy
Use of parallel computing platforms (100-500 nodes) in the processing and interpretation of large-scale datasets From November 2003 to December 2005 (2 years 2 months) Research Assistant @ From 1999 to 2003 (4 years) Intern Scientist @ Development of a Hessian preconditioner algorithm for 3-D inverse problems using approximate adjoint methods. From May 2001 to September 2001 (5 months) Albuquerque, New Mexico Area
PhD, Engineering (Computational Electromagnetics & Geophysics) @ University of Wisconsin-Madison From 1997 to 2002 M.S., Geophysics @ University of Wisconsin-Madison From 1997 to 1998 Bachelors @ Ohio University From 1993 to 1996 Michael Tompkins is skilled in: Numerical Analysis, Inversion, Inverse Problems, Modeling, Algorithms, Optimization, Scientific Computing, Statistics, Data Analysis, Signal Processing, Parallel Computing, Python, Statistical Inference, Geophysics, Electromagnetics, Regression Analysis