Staff Data Scientist, Financial Forecasting

Our mission at Duolingo is to develop the best education in the world and make it universally available. It’s a big mission, and that’s where you come in!

At Duolingo, you’ll join a team that cares about educating our users, experimenting with big ideas, making fact-based decisions, and finding innovative solutions to complex problems. You’ll have limitless learning opportunities and daily collaborations with world-class minds — while doing work that’s both meaningful and fun.

Join our life-changing mission to develop education for our half a billion (and growing!) learners around the world.

Read our blog to learn more.

About the role

Duolingo is seeking a Staff Data Scientist to own and evolve our most critical financial forecasting systems. This is a high-stakes, high-visibility role at the center of our business strategy, where experimentation, subscription dynamics, and product development converge. You will work closely with Monetization product and engineering leadership, the finance team, and senior company executives to produce forecasts that are not only accurate but explainable, trusted, and decision-ready.

We are looking for someone who deeply understands the statistical and strategic challenges of modeling constantly evolving consumer subscription products. You will need to separate signal from noise, lead investigations when reality diverges from prediction, clearly communicate the business and product impacts of modeling outcomes, and give senior leaders confidence in your numbers when the stakes are high.

:brain: You will...

• Lead the financial forecasting system. Drive methodology, implementation, maintenance, and delivery as the technical owner of our revenue forecasting stack, in close partnership with the finance team. Publish forecasts that serve as the source of truth for company-wide planning and investor-facing decisions.


• Model subscription + experimentation dynamics. Develop modeling architectures that account for experimental lift, cohort behavior, seasonality, and more, making it clear what’s real impact and what’s noise across the hundreds of experiments Duolingo runs each quarter.


• Investigate deviations and lead with insight. When forecasts miss, lead the root-cause analysis by quantifying drivers of over- or under-performance (e.g. product changes, user behavior shifts, macro trends), collaborating across functions, and clearly communicating what happened and why to senior leadership. Build and be responsible for the statistics, tools, and dashboards that scale these insights across the company.


• Advance the technical frontier. Push our evolution beyond traditional time-series extrapolation and toward more sophisticated approaches such as cohort-based forecasting, retention chaining, and/or causal impact modeling, grounded in business realism and model interpretability.


• Mentor and influence. Serve as a technical mentor to data scientists across the organization, set a high standard for modeling, communication, and statistical rigor, and develop roadmaps for forecasting infrastructure and innovation.

:check: You have...

• Forecasting domain expertise. 6+ years of proven experience building production-grade forecasting systems, with deep applied knowledge of subscription models and understanding of churn dynamics, cohort behaviors, and retention curves.


• Advanced technical and statistical skills. Advanced degree (Masters or PhD) in Data Science, Economics, Statistics, or a related quantitative field with experience assessing and applying theoretical approaches in real-world situations (ex. when to use Bayesian structural models, when causal inference is vital, when a robust heuristic is better than over-fitting noise, etc.).


• Executive communication. Experience distilling uncertainty, nuance, and model tradeoffs into streamlined, actionable narratives for finance and senior leadership partners. Ability to communicate complex forecasting logic clearly and stand by your work in moments of ambiguity.


• Collaborative style. Track record of collaboration across technical and non-technical teams, with a consistent record of integrating feedback, aligning across functions, and operating with a strong sense of ownership and service.


• Business responsiveness. Strong judgment when balancing model complexity, interpretability, and speed of development, sometimes sacrificing exactness for interpretability. Adapt quickly to fast-paced product and business changes and make model tradeoffs accordingly.


• Code quality and reproducibility. Expertise writing clean, well-documented code in Python and SQL, and building reliable, multipurpose production pipelines.


• Experience in high-experimentation environments. Past work in organizations where A/B testing is constant, complicating training data and historical baselines.