About the role

We are seeking a sharp and pragmatic Data Scientist to help us make sense of complex, large-scale data across our research and product pipelines. In this role, you will work at the intersection of statistics, machine learning, and engineering, turning messy real-world data into insights that drive decisions on model development, user experience, and business strategy. You'll partner closely with research, product, and policy teams to ask the right questions and build the analytical foundation to answer them.

What you'll do

Design and execute analyses that inform key decisions across model development, product, and operations
Build and maintain dashboards, metrics frameworks, and reporting pipelines that give teams reliable visibility into what matters
Develop and validate statistical models to understand user behavior, model performance trends, and the impact of product changes
Partner with engineering and research teams to instrument systems and ensure the right data is being captured in the first place
Identify and investigate anomalies, surface non-obvious patterns, and translate findings into clear recommendations
Establish best practices for experimentation, including A/B testing, causal inference, and measurement methodology

What we're looking for

Strong foundation in statistics, probability, and quantitative reasoning, you're comfortable with causal inference, not just correlation
Proficiency in Python or R for analysis, and SQL for data wrangling at scale
Experience working with large, messy datasets and the judgment to know when the data is telling you something real versus when it isn't
Track record of translating analytical findings into concrete decisions or product changes, not just decks
Ability to work collaboratively across technical and non-technical stakeholders and communicate complex findings accessibly
Experience with experimentation design and a healthy skepticism about metrics that can be optimized without delivering real value

Why join us?

High-Leverage Work: The analyses you run will directly shape how we develop and deploy some of the most consequential AI systems in the world
Interesting Data: The problems here are genuinely novel, there's no playbook for measuring what matters in frontier AI
Cross-Functional Reach: Work across research, product, safety, and policy, with visibility into the full arc of how decisions get made
Strong Team: Collaborate with researchers and engineers who hold the bar high and will push your thinking
Competitive Compensation: Top-of-market salary, equity, and benefits
Location and work arrangement San Francisco. This role is hybrid, with regular in-person presence expected to support close collaboration with research and product teams

Cobalt jobs

Data Scientist

Never Miss a New Opportunity