Lead Data Scientist

About the role

The Data Science Lead at Bloomerang will own the intelligence layer on top of the Unified Data Foundation (UDF), setting the technical direction for data science across the Bloomerang Giving Platform. This is a hands-on, builder role.

Responsibilities

• Prove what works, not just what correlates. Design and run the experimentation engine—randomized holdouts, uplift measurement, significance and power—to claim a fundraising action caused a lift in retention or giving, not that it happened alongside one.
• Build the predictive and forecasting models that drive donor lifetime value, retention, lapse risk, and "will we hit our goal?" forecasting—calibrated, explainable, and honest about uncertainty rather than falsely precise.
• Own model quality and evaluation. Stand up the evals, accuracy bars, and monitoring that keep our AI products and agents trustworthy—because a confident wrong answer costs a fundraiser more than no answer at all.
• Get models to production and keep them healthy. Own the ML lifecycle on Databricks and MLflow—training, deployment, versioning, and drift and performance monitoring—so models keep earning trust long after launch.
• Set the technical direction for data science. Define how we model, measure, and validate; make the call on methods and tooling; and raise the rigor bar through the quality of your own work.
• Partner across the stack. Work daily with the data engineers building the data lakehouse, the AI engineers shipping the products. Use AI tools (Claude Code, Cursor, or similar) daily for analysis, modeling, evaluation, and problem-solving.

Requirements

• Applied data science experience: 8+ years building data science and machine learning that shipped to production and moved a real metric—not models that stalled in a notebook.
• Causal inference and experimentation: deep, hands-on work with A/B testing, randomized holdouts, uplift and treatment-effect modeling, and significance and power analysis. You know why measuring impact against KPIs without a control group is the most common way to learn the wrong lesson.
• Predictive and statistical modeling: propensity, churn and retention, lifetime value, time-series and forecasting, and calibration—with the judgment to reach for the simplest model that works.
• Strong Python and SQL, and fluency with the modern ML and statistics stack (e.g., scikit-learn, gradient boosting, and the tooling behind experiment design).
• Production ML sensibility: real experience deploying, versioning, and monitoring models (Langfuse, MLflow or similar). You own outcomes after the model ships, including drift and degradation.
• Modern data platform fluency: comfortable working on a lakehouse (Databricks preferred) and partnering closely on the data models your features depend on.
• AI-Native Mindset: Hands-on AI tool usage: you already use Claude Code, Cursor, or similar AI development environments as a daily part of how you build. You can speak to where they accelerate your work and where they don't. Curiosity about the frontier: you're energized by the pace of AI-driven change—including LLM and agent evaluation—and you bring that energy into the team.
• Leadership & Ownership: Technical leadership without the org chart: you set direction through the clarity and rigor of your work, your standards, and your influence. This is a principal-level individual-contributor seat, not a people-management one. Quality-first instincts: you build evaluation, monitoring, and honest uncertainty in from day one. You'd rather ship a calibrated "we're not sure yet" than a confident answer that's wrong. Cross-functional partnership: a track record of working well with data engineers, ML and AI engineers, and product. Security and data trust: our customers trust us with their donors' data. You take that—and the consent posture behind any cross-organization analytics—seriously.

Qualifications

• Background in nonprofit, fundraising, or CRM data.
• Causal and experimentation work at product scale (experimentation platforms, sequential testing).
• LLM and agent evaluation frameworks and techniques.
• Familiarity with Data Vault 2.0 or medallion lakehouse modeling.

Skills

• Technical Depth: Applied data science experience, causal inference and experimentation, predictive and statistical modeling, strong Python and SQL, modern ML and statistics stack, production ML sensibility, modern data platform fluency, AI-Native Mindset, Leadership & Ownership, Qualifications.

Benefits

Health + Wellness: Access to generous health, vision, and dental insurance options as well as HealthiestYou, a healthcare service that offers convenient, confidential access to quality doctors 24/7, anytime, anywhere.

Time Off: Competitive PTO package that includes 20 PTO days, 3 flex days, 4 optional volunteer days, 12 paid holidays, as well as paid parental leave.

401k: Eligible for a 401k match to help invest in your future.

Equipment: Everything you need to be successful, shipped right to your door.

Pay

$138,100 - $230,200

Schedule

Permanent, full-time, fully remote position (within the U.S. and select Canadian Provinces only).