Lead Data Scientist

About the role

The Lead Data Scientist will expand the Data & Analytics Platform team's mandate to include Generative AI Analytics and tackle Snorkel's highest-leverage data science problems. They will own Generative AI Analytics, protect quality, optimize marketplace search, ranking, and recommendation models, improve forecasting, identify leverage, and advise leadership.

What You'll Do

• Own Generative AI Analytics: Build and maintain end-to-end infra (semantic layer, LLM tooling, evals, and agents) to analytically empower every team at Snorkel.
• Protect Quality: Work with Engineering and the Fraud Operations Lead to build, unify inputs for, and deploy fraud detection and contributor quality models.
• Optimize Marketplace: Architect search, ranking, and recommendation models to match project needs for Expert Contributors across domains and geographies.
• Improve Forecasting: Build predictive models to help Strategy & Operations and Finance elevate forecast accuracy with signals from across the business.
• Identify Leverage: Surface the next high-value data science opportunities at Snorkel and build the case for their prioritization with the Head of Data.
• Advise Leadership: Be a trusted thought partner to the Head of Data and leaders of other teams on data architecture, applied data science, and AI.

Qualifications

• 5+ years: data science or related experience, with a track record of shipping forecasting, ranking, recommendation, or similar models into production.
• Tech Stack: SQL, Snowflake or a similar data warehouse, and Python experience.
• Modern AI Tools: Experience deploying modern AI tools (semantic layers, LLMs, evaluation frameworks) reliably into production.
• Engineering Collaboration: Proven ability to partner with Engineering teams to bridge the gap from prototype to production system.
• Agency: Strong track record of impact without a large team or detailed roadmap.
• 0-to-1 Mindset: Comfortable building foundational systems from scratch in environments where data infrastructure is still maturing.
• Partnership: Genuinely values engaging technical and operational stakeholders to fully understand a problem and build data-driven solutions.

Bonus Points

• Experience with Streamlit, Snowflake Cortex, AI/data labeling, A/B testing, and two-sided marketplace or data product business models.