Note: The job is a remote job and is open to candidates in USA. Fusemachines is a leading AI strategy, talent, and education services provider, focused on democratizing AI. The Data Scientist role involves building and validating models for audience segmentation across various media platforms, collaborating with ML Engineers and product teams to enhance audience capabilities.
Responsibilities
- Build statistical and ML models to create, expand, and score audience segments from survey, panel, purchase, and media-exposure data
- Develop propensity and lookalike models that scale small seed audiences to addressable populations
- Lead data fusion work, combining deterministic and probabilistic sources into one representative consumer view while correcting for bias
- Ideate on making the product better: propose new audiences, features, and methods, and prioritize them with the US team
- Own audience measurement analytics: reach, overlap, index strength, and incremental lift (A/B, matched control, causal inference)
- Partner with ML Engineering to move models into reproducible, monitored production, under privacy-by-design principles
Skills
- ~5+ years in applied data science, with models that reached production or client delivery
- Sound statistical modeling (survey stats, causal inference, experimental design, propensity/uplift)
- Working knowledge of Databricks; comfortable in a Spark / Snowflake environment
- Strong Python (pandas, scikit-learn, statsmodels) and SQL against large databases
- Degree in a quantitative field (Statistics, Data Science, Economics, CS, Math, or similar)
- Media / advertising background (strong plus)
- Audience/identity or ad tech exposure (DSP/SSP, DMP/CDP, clean rooms, identity graphs)
- Familiarity with privacy-preserving methods and GDPR/CCPA in data collaboration
Company Overview
Company H1B Sponsorship