Bayesian Econometrician

in4mation insights
Needham (just outside Boston) - United States
S: Competitive

In4mation Insights is an internationally recognized firm in the marketing research and analytics field and is known for its industry-leading theoretical and practical leadership and advanced technology. The firm's leaders each have more than 30 years' experience in consumer research, data analytics, modeling, and technology development to deliver answers related to Pricing, Marketing ROI, New Product Development, Market Structure and Consumer & Market Segmentation. The firm has a worked with a number of Fortune 200 companies; our experience includes industries such as consumer goods, IT, Telecom, Financial Services, and Media.

We have developed a suite of hierarchical Bayesian econometric models for use on large datasets using high performance parallel computation. We need talented analysts and developers who understand computation and Bayesian statistics to continue to enhance the existing code base and to develop new models as needs arise. These latter models include data mining tools, Bayesian churn models, and Bayesian variable selection methods. We typically use R, WinBUGS, or Python for prototype development and then transfer the code to C++ for speed and scalability when necessary.

You would be working as part of our software development team and our marketing science team, along with our Board of Science advisors, some of whom include John Liechty (Penn State), Michel Wedel (Maryland), Greg Allenby (Ohio State), and Pete Fader and Eric Bradlow (Wharton). Please investigate these advisors; they are world-class.
We have our own in-house high performance computation cluster with over 200 cores and 2 TB of memory running Linux.

In addition to statistical software and model development, the position involves advanced modeling of large transactional databases and consumer surveys as part of client engagements. Consultation with clients and other staff on interpretation and analysis of advanced models is also a key part of this job.

Our clients are Fortune 200 companies who trust us with their most difficult and challenging analytics problems in marketing and operations.

The position

For Bayesian statistics, the candidate MUST know:
• MCMC methods, including Gibbs and Metropolis-Hastings sampling. Exposure to Slice, Hamiltonian, and other samplers would be great.
• Econometrics and/or mathematical statistics, with knowledge of regression and choice models (e.g. logit, probit). Experience with Vector Auto Regression (VAR), state space, UCMs, etc is a plus.

The candidate should:

• Have excellent skills with R
• Experience with Python or a similar language
• Experience with or exposure to SAS
• Skill and experience with C++ is a plus

Compensation
• Competitive salary with potential for bonus
• Strong individual contributors can become stockholders in the company
• 100% individual health and dental insurance, 75% for family
• Disability & life insurance policies
• 401K plan with employer matching
• Profit sharing plan
• PTO bank – combination of vacation, sick, and personal days

Job summary