Director, Data Science (Finance)

Datasite•Minneapolis, MN

About The Position

As the Director of Data Science, you will be the founding architect of our data science function and you will lead the evolution of our intelligence layer. You are responsible for transforming complex, abstract business problems into rigorous predictive models and experimentation frameworks. You will lead 1 or 2 Data Scientists to build "Data Intelligence Products" example: automated forecasting tools, risk prediction engines, and ML-driven initiatives that directly impact revenue and profitability. You are part Hands-on Scientist and part Strategic Builder, bridging the gap between abstract problem solving and concrete business outcomes.

Requirements

10+ years of experience in Data Science or Advanced Analytics, with 3+ years in a leadership capacity.
Proven track record of building and deploying predictive models (Revenue, Risk, or Forecasting) that achieved high business adoption.
Modern Stack Experience: Deep familiarity with the Snowflake + dbt + Power BI ecosystem is a significant advantage.
Proven track record of building and deploying predictive models using Python and SQL that achieved high business adoption.
Toolkit Expertise: Hands-on experience with ML orchestration tools and automated testing for model performance (e.g., monitoring for Data Drift and Model Decay).
Cloud Infrastructure: Experience with cloud-based ML platforms (e.g., Azure ML, AWS SageMaker, or Databricks) and how they integrate with data warehouses like Snowflake.
The ML Toolbelt: Deep expertise in the Python Data Science stack (e.g., scikit-learn, XGBoost, LightGBM) and deep learning frameworks (e.g., PyTorch or TensorFlow).
Predictive Expertise: Deep experience in time-series forecasting, supervised learning, and causal inference.
Time-Series & Forecasting: Mastery of libraries dedicated to financial and demand forecasting, such as Prophet, statsmodels, or sktime.
MLOps & Deployment: Experience with model lifecycle management tools (e.g., MLflow, Weights & Biases) and deploying models via containers (Docker/Kubernetes) or as serverless functions.
Statistical Logic: You don't just run models; you understand the "why" behind the math and can defend your methodology to technical and non-technical audiences.
Mathematical Depth: Deep expertise in supervised/unsupervised learning, Bayesian statistics, time-series analysis, and causal inference.
Generative AI & LLMs: Working knowledge of integrating LLMs (via LangChain, OpenAI API, or Hugging Face) into business workflows for unstructured data analysis.
The Modern Data Stack: Proficiency in using Snowflake as a feature store and dbt for feature engineering.
Finance Acumen: You understand the levers of a P&L and how predictive modeling impacts revenue and margin.
Communication: Exceptional ability to simplify complex "black box" concepts for executive stakeholders.

Nice To Haves

Foundational Builder Experience: Experience hiring, mentoring, or designing a data science function is a significant plus.
An MBA combined with a technical background is highly valued
Entrepreneurial Spirit: You are excited by the prospect of building a department from the ground up, from selecting tools to hiring the team.

Responsibilities

Intelligence & Model Development (The Engine) High-Impact Modeling: Directly oversee and contribute to the development of predictive models for revenue forecasting, profitability, and demand planning.
Risk Prediction Tools: Architect and deploy tools for predictive financial risk assessment, helping the business identify and mitigate volatility before it occurs.
ML/AI Roadmap: Define the vision for how AI/ML will be integrated into our modern data stack (Snowflake/dbt/Power BI) to automate complex decision-making.
Experimentation Rigor: Establish the framework for A/B testing and statistical experimentation to validate business strategies and product changes.
Strategic Leadership & Ambiguity Management (The Bridge) Abstract Problem Solving: Serve as the primary partner to the C-suite, translating vague business challenges into structured data science projects with clear ROI.
Stakeholder Management: Work cross-functionally (Finance, Marketing, Ops) to ensure that predictive insights are not just "interesting," but are integrated into the operational workflow.
Data Productization: Partner with Data Engineering to ensure models are "production-ready," moving them from local scripts to automated, reliable outputs in Power BI.
Department Building & Talent Pipeline (The Growth) Team Scaling: Act as a "Player-Coach" to the current Data Science team while identifying the specific skill gaps (e.g., NLP, Deep Learning, MLOps) needed for future hires.
Talent Pipeline: Proactively build a network and recruitment strategy for future Data Analytics and Data Science roles to ensure rapid scaling as the function proves its value.
Standard Setting: Establish the "Data Science Playbook"—defining our standards for code quality, model validation, and documentation.