Machine Learning Data Scientist, Forecasting Platform
- Location
- San Francisco, CA
- Work model
- Hybrid
- Level
- Senior
- Salary
- $270k/yr
- Posted
Skills
About this role
About Anthropic
Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.
About the role
Anthropic's business changes quickly: new models, products and pricing land every few weeks, and each one can move the forecast. The Forecast Science team is hiring a machine learning data scientist to build and own the forecast platform behind the revenue forecasts. You'll build the backtesting and accuracy discipline that makes our forecasts trustworthy, and scale the systems that let the team develop and run our statistical models faster. We expect candidates to lead with depth in machine learning engineering, in addition to a strong statistics background. This is a build role on a small team: you'll write production code and keep your systems running. The work is technical, fast-moving and highly visible, and it spans Finance, Data Science and Data Engineering. You have a bias for action, choose practical solutions over complex ones, and want to learn how finance works. We're looking for someone who wants to build forecast systems that learn from every cycle, to accelerate how we predict our business and influence broader strategic decisions.
Key responsibilities
• Build and own the forecasting platform end to end, from data and training to backtesting, deployment and monitoring, with results that planning teams can self-serve
• Run forecast models in production with tested, versioned, reviewed code that anyone on the team can rerun, and make it fast and safe to test a new model, from classical methods to foundation models
• Automate backtesting, accuracy tracking, drift monitoring and input data checks for every model, and use the results to improve quality cycle over cycle
• Design and scale the cloud systems that run the forecast, including scheduled pipelines, parallel training and backtests, monitoring and compute cost
• Own the infrastructure for models and adjustments, so changes are fast and traceable
• Build agentic workflows with Claude and self-serve tools that speed up modeling, adjustments and what-ifs, with checks that keep the numbers trustworthy
• Partner with Finance, data science and data engineering on what to build next, and explain trade-offs in plain language
Minimum qualifications
• Have substantial experience in machine learning engineering, data science, forecasting or quantitative finance, including shipping and maintaining machine learning models that others rely on
• Are deeply fluent in Python and SQL, and can self-serve in a cloud warehouse
• Have designed and scaled data or ML systems in a cloud environment (such as AWS or GCP), including orchestration, parallel compute, monitoring and cost
• Have a strong statistics foundation, with hands-on experience running time-series methods (Prophet, ETS, ARIMA, gradient-boosted approaches, neural forecasting, hierarchical reconciliation), and use it to measure uncertainty and bias and tell a real gain from noise
• Have built backtesting and accuracy tracking that catch production failures (leakage, drift, upstream data changes), and are comfortable having your models scored openly
Preferred qualifications
• 6+ years of relevant industry experience (machine learning or applied data science) and a Master’s degree or PhD in a quantitative field
• Depth in causal inference (difference-in-differences, synthetic control, Bayesian structural time series, event studies)
• Experience with