About

I am an economist/data scientist with 9+ years of experience in causal inference, experiment design, and statistical modeling. Currently applying advanced Bayesian methods in oncology drug development at AstraZeneca. Expert in building hierarchical state-space models, survival analysis, and clinical trial analytics using R, Stan, and Julia. Experienced in using Bayesian networks for causal identification in experimental and non-experimental contexts. Proficient in reinforcement learning for partially observable environments. Focused on rigorous quantification of uncertainty for decision-making.

📄 Download Resume (PDF)

Skills

Programming

Expert: R, Tidyverse, Stan, Julia, Flux.jl, Turing.jl
Proficient: C/C++, Ruby, Python, SQL

Statistical Methods

Advanced Bayesian Methods: Hierarchical modeling, Gaussian processes, state-space models, survival analysis
Causal Inference: Experimental design, instrumental variables, difference-in-differences, regression discontinuity
Machine Learning: Deep reinforcement learning, particle filters, Monte Carlo methods, Bayesian neural networks

Tools

RStudio, VS Code, AWS EC2/S3/ECR, Azure, Docker, Cluster Computing (SLURM), Databricks, Git and Github, Snowflake, PostgreSQL, cmdstanr, targets workflow management, Domino, agentic AI (Claude, ChatGPT)

Education

Boston University
Ph.D. in Economics

Boston, MA
Granted 2014

Focus: Development economics, health economics, and experimental economics

The American University in Cairo
B.S. in Computer Science
Minor: Mathematics

Cairo, Egypt
Granted 1999

Experience

Senior Data Scientist - AstraZeneca

Aug 2023 - present

Advanced Bayesian State-Space Modeling for Oncology Clinical Trials

• Developed sophisticated hierarchical Bayesian state-space models for tumor growth dynamics using the Stein-Fojo framework with simultaneous regression and growth components
• Implemented log-transformed state-space formulation with Gaussian process temporal correlations for irregular time intervals
• Built comprehensive Stan models with efficient MCMC sampling across population, trial, and patient levels
• Created scalable computational framework for patient-level state calculations
• Developed RECIST response prediction models for clinical decision support
• Integrated survival analysis with competing risks for progression-free survival and overall survival endpoints
• Implemented real-time forecasting capabilities for censored patients
• Created comprehensive validation framework with leave-future-out cross-validation for temporal model validation

Independent Researcher

Nov 2022 - Aug 2023

Reinforcement Learning for Intervention Optimization (2022-2023)

• Designed and built a Julia package for a reinforcement learning solution (partially observable Markov decision process), using Monte Carlo tree search, to optimize intervention implementation and evaluation decisions
• Used Bayesian hierarchical models to predict future states in POMDP
• Used particle filters for faster online planning
• Project writeup: The Funder’s Meta-Problem
• Implemented various deep reinforcement learning agents and models using Julia and Flux.jl

Senior Data Scientist - Opendoor

May 2021 - May 2022

• Designed and built continuous randomized experiments to estimate and predict demand and supply elasticity in real estate markets. These were the principal models used in determining offers for home buying/selling.
• Used Gaussian processes to construct nonlinear time-series prediction models.
• Used hierarchical models to capture heterogeneity in different markets, offer types, etc.
• Designed a survival and time-series model to capture the duration homes are on the market and how this affects elasticity.
• Solicited stakeholders to capture prior beliefs about both the structure of models and distributional uncertainty, which is critical in facilitating decision-making in small data environments.
• Designed non-experimental causal identification strategies for contexts where experiment manipulation is not possible.

Independent Researcher/Consultant

May 2019 - May 2021

• Built a structural Bayesian model representing a behavior model, using data from a large-scale social experiment conducted in Kenya.
• Worked on authoring an economics research paper on findings from the experiment.
• Built a hierarchical model to evaluate the effectiveness of a scaled economic intervention program in Bangladesh.
• Used Gaussian processes to model correlation in intervention effects based on spatial proximity.
• Designed and built a hierarchical model investigating the heterogeneity of COVID-19 transmission in different countries under different stay-at-home restrictions.
• Used a simulated method of moments approach to fit the AcceleratingHT model’s parameters to the CGD model’s predictions of vaccine success.
• Implemented R package for bounded identification of counterfactuals, in directed acyclic graphs, in the presence of imperfect compliance with experimental treatment assignment.

Economist - Evidence Action

May 2014 - May 2019

• Designed and oversaw the implementation of two large-scale experiments in Kenya and Bangladesh, evaluating interventions aimed at altering health behavior and incentivizing seasonal migration to fight malnutrition, respectively.
• Responsible for evaluating any evidence-based decisions by the organization.

Software Design Engineer in Test - Microsoft

Apr 1999 - Mar 2007

• Constructed testing framework for the Windows Debugging Tools and command line utilities.

Research

Publications

• Nathan Barker, C. Austin Davis, Paula López-Peña, Harrison Mitchell, Ahmed Mushfiq Mobarak, Karim Naguib, Maira Reimão, Ashish Shenoy, and Corey Vernot, “Migration and Resilience during a Global Crisis.” Accepted: European Economic Review.

Working Papers

• Harrison Mitchell, A. Mushfiq Mobarak, Karim Naguib, Maira Reimão, and Ashish Shenoy, “External Validity and Implementation at Scale: Evidence from a Migration Loan Program in Bangladesh.”
• Edward Jee, Anne Karing, and Karim Naguib, “Optimal Incentives in the Presence of Social Norms: Experimental Evidence from Kenya.”