At Neighbor, we’re building the largest hyperlocal marketplace the world has ever seen. We’ve raised over $75 million from top-tier investors such as Andreessen Horowitz and the CEOs of DoorDash, StockX, and Uber. Our marketplace is already flourishing in all 50 states and we’re just getting started!
We're excited to add a Data Scientist Intern to our Data & Analytics team. You'll work on the machine learning models we already run in production, including lifetime value, unit economics, and forecasting, making them more accurate and showing with evidence that they've improved. You'll also help Product, Marketing, and Sales divisions design A/B tests and interpret the results. Your work feeds directly into decisions across a marketplace that operates in nearly every U.S. city. This is a great fit for a PhD student who wants to apply their research skills to live business problems. You'll report to our Data & Analytics manager, with regular code review and hands-on mentorship.
Our stack: Python, dbt, and Dagster on Redshift and Athena, with a Cube semantic layer and Superset for BI.
The Problems You'll Solve
Improve our lifetime value and unit economics models. Retrain them, re-engineer their features, and validate their predictions against realized outcomes.
Audit inherited models: find the leakage, the stale hardcoded assumption, the segment where performance quietly falls apart, and the feature that's doing less work than everyone believes.
Build forecasts our operators actually plan against: demand and supply by market, revenue, and the levers that move them.
Partner with Product, Marketing, and Sales to design tests before they launch. Catching an underpowered test in the design review is worth more than any analysis you can do afterward.
Analyze results and make a call and be candid about what each design can and can't identify
Become a subject matter expert on Neighbor's product, users, and marketing life cycle. The modeling is the easy part; knowing which features mean something is the hard part.
Qualifications
PhD-level candidate currently enrolled in a quantitative field (Computer Science, Statistics, Math, Physics, Data Science, etc.) completing your PhD by May of 2028; transcript required
Demonstrated research, coursework, or project experience applying statistics and machine learning to real-world or complex datasets
Strong SQL and fluent Python for modeling are required; experience with dbt or semantic layers like Cube is a big plus, but we're happy to train you on our specific transformation stack
Academic or project-based experience with A/B testing design or statistical hypothesis testing
Real grounding in inference, not just fitting: you can explain what your confidence interval means, why the p-value moved when you added a second metric, how leakage sneaks into a validation split, and when you don't have the data to answer the question
Strong problem-solving mindset and eagerness to diagnose and improve existing code and statistical models
Intellectual stubbornness in a productive direction: you keep pulling on a thread when the numbers don't reconcile, and you don't ship an explanation you don't believe
Clear communication with non-technical stakeholders: you can explain a result and its uncertainty without either overclaiming or hiding behind jargon
