Data Science, Finance & Strategy

About the role

Anthropic’s Finance Analytics & Business Intelligence team is hiring a senior individual contributor to own how we measure the value of our models and our position in the market. These are open questions without an established playbook: how much value do our models deliver per dollar and per token, how is that changing with every launch, how do we compare to the rest of the frontier?

You’ll own how Finance quantifies relative model value and market position: maturing our cross-product benchmark suite, building task-cost and price-elasticity estimates that inform live pricing and packaging decisions, sourcing and running capability and market analysis around every model launch, and standing up forecasting on third-party and survey data. The work is open-ended and technical, and you’ll operate as an analytical lead, partnering closely with Product Finance and our model performance Data Science teams.

Key responsibilities

• Build the relative-value measurement system: evolve our cross-product benchmark into a durable, trusted read on model and product value, spanning coding, agentic, and product-shaped tasks
• Inform pricing and packaging: construct task-cost approximations and price-elasticity estimates across differently priced products, and carry them into decisions
• Own launch and market analytics: run analytics around model launches, including capability-based revenue analyses and views of the broader market
• Deepen our market understanding: evaluate and integrate external datasets and research to strengthen our read on the market and how it's evolving
• Partner with Product Finance: take open-ended pricing, packaging, and positioning questions from vague ask to decision-grade answer
• Raise the bar: land narratives in executive forums and uplevel the team’s product-finance analytics practice by example

Minimum qualifications

• Put shape around ambiguity: you’ve personally defined the measurement approach for questions nobody knew how to answer, without waiting for a fully specified ask
• Land narratives with executives: your analyses have changed pricing, product, or competitive decisions, and you can simplify for senior leaders without losing rigor
• Stay hands-on at senior scope: you still write the SQL and Python yourself, and you’d rather ship a defensible v1 with honest error bars than wait for perfect data
• Are inherently curious: you go one level deeper than asked and are energized by how fast models, products, and the market are moving
• Thrive amid shifting priorities: you juggle multiple fast-moving workstreams and stay effective when the plan changes weekly
• Work fluently with modern tooling: you’re strong at data visualization, use Claude and AI tools as force multipliers in analysis and BI, and can self-serve your own workflows across SQL, Python, dbt, and a cloud warehouse

Preferred qualifications

• Experience designing evals or benchmarks for AI models or products
• Pricing and packaging analytics at scale, including elasticity estimation
• Market share estimation from imperfect third-party, panel, or survey data
• Fluency in the LLM model and product landscape
• Dimensional modeling and warehouse design experience (grain, SCDs, point-in-time correctness)
• Cloud platform experience (AWS, GCP) with orchestration, CI/CD for data, and testing/observability