Research Consultant – Bayesian Modeler

Hiring: Research Consultant – Bayesian Modeler Location: Remote (Must be based in the US) Duration: Immediate start through March 15, 2027, with possible extension Hours: 20–35 hours per week Eligibility: Must be authorized to work in the US without sponsorship

About the Role

We are seeking an experienced Bayesian Modeler to support a research project at Tufts University's Digital Planet, focused on developing a computational model of AI readiness across 27 low- and middle-income countries. The consultant will work on Bayesian modeling, statistical analysis, model validation, and developing reproducible research pipelines.

Key Responsibilities

Develop, estimate, and validate Bayesian hierarchical and latent-variable models. Build and test measurement, index, or population segmentation models. Compare alternative statistical models and perform sensitivity analysis. Analyze multi-country and multi-year panel data, including missing data. Develop reproducible Python pipelines and document methodologies.

Required Qualifications

Hands-on experience with Bayesian hierarchical or latent-variable modeling. Experience building comparable models end to end, including developing or validating weighting schemes. Experience with multi-country or multi-year panel data and incomplete datasets. Strong Python skills (pandas, NumPy, SciPy) and experience with PyMC, NumPyro, or Stan. Familiarity with Git or other version-controlled repositories. Strong analytical, research, and written communication skills.

Preferred Qualifications

Experience with composite indicators, cross-country measurement models, or population segmentation. Familiarity with household survey data and digital access research is a plus.