Senior Data Scientist

Returnalyze is the first AI platform that identifies and prevents retail returns before they happen. Where traditional BI tools only report what happened, our platform finds patterns across billions of variables to predict and prevent returns, delivering clear recommendations with quantified revenue opportunities. We’re trusted by leading retailers including J.Crew, Abercrombie & Fitch, GAP, Nordstrom, Brooks Running, Merrell, and Costco.

We’re hiring a Senior Data Scientist to push the analytical frontier of the platform. Someone with the statistical depth and engineering range to invent new insight types, sharpen our predictive models, deepen our understanding of customer behavior, and expand what the platform can detect and recommend. This person owns analytics direction: not just building what’s asked but defining what should be built and what “good” looks like.

What You'll Do

• Design new methods for anomaly detection, forecasting, root-cause attribution, and causal impact measurement.
• Ship them into production across dozens of client environments and billions of transactions.
• Push our predictive modeling forward and flag high-risk products earlier in their lifecycle.
• Build customer-level analytics: lifetime value, acquisition strategy, and high-risk customer identification.
• Connect return behavior to customer value, so clients can act on who they acquire and retain.
• Improve estimation on noisy, heterogeneous data and quantify uncertainty rigorously.
• Invent and prototype new insight types across retail functions such as eCommerce, product design, merchandising, operations.
• Deepen AI/LLM use in the product (including Gemini).
• Establish whether a recommended action changed outcomes.
• Own work end to end: method design, production code, testing, and validation.

Qualifications

• Statistics & data science depth. Judgment to choose and defend methods on messy, large-scale data. Bayesian methods, forecasting, and causal/quasi-experimental inference highly valued.
• Production coding ability. Production coding ability is a must, preferably Python and SQL on large datasets. Must be able to write and validate production-quality code, not just theory: enough engineering depth to ensure outputs and implementation make sense, and technically strong enough to direct others and judge whether the math, outputs, and code are correct. A track record of shipping unattended production systems.
• Machine learning / forecasting. Predictive modeling that flags risk earlier in the product lifecycle.
• AI/LLM implementation. Real experience integrating LLMs into analytics applications; high-capacity use of Claude Code or similar AI coding tools.
• Delivery instinct. Rapid, iterative shipping. A useful first version, measured, then improved and the initiative to find where new analytics create the most client value.
• Able to be in the Burlington, MA, office 3 days per week (Tues, Weds, Thurs).
• 5-7 years of relative experience.

Strongly Prefered

• Retail experience, especially merchandising or product assortment.
• Experience with Snowflake or Big Query at scale.
• Data engineering fluency with large datasets.

Behaviors

• A startup mindset: scrappy, adaptable, and energized by ambiguity, always finding a way to make things happen.
• Passionate about using client data to reshape business outcomes and disrupt the status quo.
• Strong problem-solving skills.
• A collaborative approach. You work well with team members in the office and remotely.