Statistical computing is unlike other forms of software development, even other forms of scientific software development. It is challenging because it requires integrating a variety of skills and technologies. It can also demand a great deal of resources, though these resources can be reduced by using appropriate algorithms and infrastructure.
Although statistical computing is unique, it cannot be isolated. Statistical problems do not exist in a vacuum. They arise in some larger context, and so their solution has to fit into that context. Statistical software must draw data from somewhere and feed conclusions somewhere, and so statistical software must integrate with other software systems.
Years of experience in scientific and statistical computing have prepared me to help companies with these challenges. I studied differential equations and numerical computing while completing a PhD at The University of Texas. After a postdoc at Vanderbilt University, I worked in industry doing digital signal processing and general software development. I then spent over a decade at MD Anderson Cancer Center managing statistical software development.
|“John helped us with a wealth of statistical programming knowledge, plus a great understanding of what constitutes good statistical practices. This helped our company’s project with NIH.” — Marc Abrams, CTO, Harmonia|
Since leaving MD Anderson I have consulted for numerous companies, helping them with challenges in statistical computing. Clients have included software giants, pharmaceutical companies, law firms, and companies from a wide variety of other areas.
Get expert help with your statistical computing challenges.