Data Scientist

The ideal candidate combines strong Python and SQL skills with practical judgment: they can build reliable training datasets, validate model behavior rigorously, and ship maintainable solutions without waiting for a model to be “perfect.”. Validate features and model performance beyond in-sample metrics by assessing stability across time periods and cohorts, inference-time availability, potential leakage, and real-world plausibility.

Data Scientist

Location:  West Metro of Minneapolis

We are seeking a collaborative, hands-on Data Scientist to develop, enhance, and deploy machine-learning and advanced analytics solutions. This is a broad, generalist role supporting work across risk, personalization, and other business-focused analytical initiatives.

The ideal candidate combines strong Python and SQL skills with practical judgment: they can build reliable training datasets, validate model behavior rigorously, and ship maintainable solutions without waiting for a model to be “perfect.”

Responsibilities

• Enhance, maintain, validate, and monitor existing machine-learning models, representing approximately 80% of the role.

• Design and develop new predictive models and advanced analytics solutions for business needs, representing approximately 20% of the role.

• Perform exploratory data analysis to identify opportunities, assess data quality, and guide modeling approaches.

• Independently engineer training datasets, including target definition, sampling strategy, class-balance considerations, and temporal data integrity.

• Select and evaluate appropriate modeling approaches based on the business problem, available data, explainability needs, inference requirements, and operational complexity.

• Validate features and model performance beyond in-sample metrics by assessing stability across time periods and cohorts, inference-time availability, potential leakage, and real-world plausibility.

• Use time-aware validation approaches when working with temporal data and investigate distribution shift, seasonality, and performance degradation.

• Take models from exploratory development through production deployment using the team’s established, low-overhead deployment process.

• Refactor notebook-based work into maintainable production code with version control, testing, and appropriate deployment practices.

• Contribute to development and release processes across DEV, QA, and PROD environments.

• Partner closely with technical and business stakeholders to clarify requirements, gather feedback, and iterate on model solutions.

• Work effectively in a two-week sprint cadence, delivering usable improvements incrementally.

• Manage multiple workstreams and maintain context across concurrent projects and stakeholder conversations.

• Create visualizations or dashboards as needed to communicate analysis and model insights.

Qualifications

• 3–5 years of professional experience in data science, machine learning, advanced analytics, or a related field.

• Strong proficiency in Python and SQL.

• Hands-on experience with Snowflake; this is strongly preferred.

• Experience developing and evaluating a range of model types, such as linear regression, classification models, collaborative filtering/recommendation approaches, ranking models, or other predictive techniques.

• Demonstrated ability to structure high-quality model training datasets and assess target sampling strategies.

• Strong understanding of:

  • Feature engineering, feature validation, and feature importance.

  • Data leakage and ensuring features are available at prediction time.

  • Temporal train/validation/test splits.

  • Distribution shift, cohort-based validation, and model generalization.

  • Class imbalance, negative sampling, and evaluation methods aligned to real-world inference scenarios.

• Experience moving machine-learning models from Jupyter notebooks into production environments.

• Working knowledge of GitHub, version control, code review practices, testing, and environment separation across DEV, QA, and PROD.

• Experience working in VS Code or comparable modern development environments.

• Ability to interpret technical blueprints or specifications, ask effective clarifying questions, and independently execute against defined requirements.

• Strong communication skills and a collaborative, iterative working style.

Preferred Qualifications

• Experience with model monitoring, containerization, and production model lifecycle practices.

• Experience supporting analytics or machine-learning use cases in personalization, risk, behavioral prediction, or related domains.

• Experience building dashboards, visualizations, or self-service analytical tools.

• A GitHub portfolio or examples of production-quality data science work.