Data Science Intern

FocusKPI Inc. • Boston, MA

17h • Remote

About The Position

We are looking for a curious and analytically minded Data Science Intern to support the development and evaluation of statistical and machine learning models for marketing measurement. This role is designed for someone with strong quantitative fundamentals who wants hands-on experience applying regression, model diagnostics, validation, and data analysis to real business problems. You will work closely with experienced data scientists, learn how modeling choices affect interpretation and business decisions, and contribute clean, reproducible analytical work.

Requirements

• Strong foundation in statistics and regression: understanding of linear regression, key model assumptions, coefficient interpretation, regularization concepts, and basic statistical inference.
• Solid quantitative fundamentals in probability, statistics, and linear algebra; familiarity with calculus or optimization concepts is helpful.
• Working knowledge of Python for data analysis and modeling, including common data-science libraries; basic to intermediate SQL skills for data extraction and transformation.
• Understanding of model evaluation: training versus validation data, cross-validation, common regression metrics, overfitting, and the importance of out-of-sample performance.
• Ability to reason through modeling problems: investigate unexpected results, form hypotheses about root causes, test alternatives, and explain conclusions using evidence.
• Clear communication skills: ability to explain analytical methods, assumptions, results, and limitations to technical teammates and learn from feedback.
• Currently pursuing a degree in statistics, computer science, data science, machine learning, applied mathematics, econometrics, operations research, or a closely related quantitative field.

Nice To Haves

• Coursework, research, or project experience using regression, time-series analysis, statistical modeling, or machine learning.
• Exposure to Marketing Mix Modeling (MMM), marketing analytics, attribution, or other measurement problems.
• Basic understanding of concepts such as adstock, saturation, incremental impact, ROI, or response curves.
• Familiarity with A/B testing, causal inference, simulation, sensitivity analysis, or confidence intervals.
• Experience with Python libraries such as pandas, NumPy, statsmodels, scikit-learn, SciPy, or similar tools.
• Previous internship, research assistantship, academic project, or independent project involving real-world data is a plus.

Responsibilities

• Support the development and evaluation of models including regression, time-series, and other statistical or machine learning approaches, with attention to predictive performance, stability, and interpretability.
• Prepare and explore data using Python and SQL; perform data-quality checks, feature construction, descriptive analysis, and visualization to understand modeling inputs and outcomes.
• Apply core model-validation techniques such as train/validation/test splits, cross-validation, baseline comparisons, and appropriate performance metrics.
• Investigate common statistical issues including multicollinearity, overfitting, residual patterns, autocorrelation, heteroskedasticity, and unstable coefficients, with guidance from senior team members.
• Test and compare reasonable modeling choices such as feature transformations, regularization settings, and model specifications, and summarize how these choices affect model results.
• Interpret model outputs and connect technical findings to practical marketing or business questions while clearly stating assumptions and limitations.
• Contribute to reproducible analytical workflows for model training, validation, sensitivity checks, and result comparison.
• Write clear Python and SQL code and communicate methods, findings, assumptions, and open questions in a structured and understandable way.

What This Job Offers

• Job Type: Full-time
• Career Level: Intern
• Education Level: No Education Listed