Data Scientist

Responsibilities

• Build, validate, and refine demand forecasting models for GTI's retail, wholesale, and other emerging business verticals across daily, weekly, monthly, and quarterly forecast horizons
• Engineer new features for the Snowflake Feature Store — drawing from retail sales history, inventory movement, weather data, customer demographics, and external signals — to improve model accuracy across store, product, market and other dimensions
• Develop and test new model candidates against GTI's established backtesting framework; interpret backtest results and surface findings to inform promotion decisions
• Investigate forecasting errors and anomalies: identify when model performance degrades, diagnose root causes (data drift, structural breaks, new store openings, regulatory changes), and propose remediation
• Conduct dimensionality reduction and principal component analysis to understand primary feature importance
• Collaborate with the Manager to evolve the feature engineering roadmap — identifying signals worth building, data gaps worth closing, and model architectures worth exploring
• Design, validate, and execute analytical studies that answer business-user’s operational questions which can then be modeled and replicated by our data analyst AI agent to further promote self-service
• Build reusable analytical frameworks on top of GTI's curated data layer (retail sales, inventory, customer, loyalty, workforce) that can be repeated, parameterized, and handed off to the business
• Contribute to quasi-experimental modeling: pre/post adult-use launch performance, store cohort comparisons, product mix attribution, and discount effectiveness
• Translate analytical findings into clear written summaries and visualizations that non-technical stakeholders can act on
• Identify patterns in the data that surface new questions worth asking — and bring those to strategy discussions with the Manager

Qualifications

• 2+ years of hands-on experience in a data science, quantitative analyst, or ML engineering role — with demonstrable work in model building, feature engineering, or statistical analysis
• Strong Python skills for data manipulation, modeling, and analysis (pandas, scikit-learn, statsmodels, or equivalent). Jupyter notebook development or equivalent experience
• Strong SQL skills — comfortable writing complex queries across multiple joined tables, aggregating at multiple grains, and debugging data quality issues in query output, while validating accuracy and trust
• Working experience with supervised and unsupervised ML methods: gradient boosting, time series models, random forest, decision trees, etc
• Ability to communicate analytical findings clearly in writing — you don't just run the analysis, you explain what it means and what to do about it
• Intellectual curiosity and a bias toward figuring things out — this role requires navigating real, messy data in a complex multi-state retail operation

Preferred

• Experience with time series forecasting methodologies (ARIMA, Prophet, LightGBM/XGBoost for tabular time series, or similar)
• Experience with advanced machine learning modeling techniques and algorithms such as Bayesian inference, Deep Learning neural networks, k-means clustering, etc
• Familiarity with feature store concepts or structured feature engineering pipelines
• Exposure to Snowflake, Snowpark, or cloud data warehouse environments
• Experience with dbt or working in a layered data warehouse (raw → refined → curated) — understanding where data comes from matters here
• Experience prototyping and productionizing data products such as Streamlit apps
• Basic familiarity with LLM-powered tooling or AI agent frameworks — not required, but exposure gives you context for where the team is headed

Additional Requirements

• Must pass any and all required background checks
• Must be and remain compliant with all legal or company regulations for working in the industry
• Must be a minimum of 21 years of age