Junior/Middle Data Scientist

You’ll join our Marketing Science team to build, run, and explain econometric & causal models that quantify the impact of media and non-media drivers on business outcomes. Day-to-day you’ll own MMM pipelines (from data ingestion to reporting), run incrementality tests, and turn results into budget/flighting recommendations that marketers can act on. 

Key responsibilities 

• MMM development & maintenance: specify and fit models (adstock, saturation, carryover; hierarchical/Bayesian variants), manage feature engineering, priors, diagnostics, and retraining cadence. 
• Causal measurement: design & analyze geo-tests and holdouts. 
• Optimization & scenarioing: deliver budget allocation and “what-if” tools; communicate trade-offs, constraints, and uncertainty. 
• Stakeholder enablement: translate model outputs for marketers, and media planners; produce concise decks/dashboards with clear recommendations. 
• Documentation & governance: maintain modeling playbooks, data dictionaries, and QA checklists; ensure reproducibility and ethical use of data. 
• Readiness to work from 15:00 till 23:00 in order to have an optimal overlap of work hours with USA based teams.

Qualifications 

• Solid statistics/econometrics foundation (regression, time series, regularization; basics of Bayesian inference). 
• Comfortable in Python (pandas, numpy, statsmodels / scikit-learn); SQL for joins/aggregations; Git. 
• Ability to clean/validate messy marketing data and communicate results in clear visuals (Plotly/Matplotlib, Tableau/Looker Studio). 
• English B2+ for async docs and client-facing calls. 

Nice to have 

• Experience with marketing/media planning concepts (reach/frequency, flighting, creative wear-out). 
• Cloud familiarity (GCP/AWS), containerization, CI. 
• Forecasting (SARIMAX/ETS), gradient boosting (XGBoost/LightGBM), uplift modeling. 
• Privacy-preserving measurement patterns. 
• Bayesian tools (PyMC/Stan) and MMM libraries (e.g., Meta’s Robyn, LightweightMMM. 
• Experience shipping schedulable pipelines (Airflow/dbt) and working in BigQuery/Snowflake/Postgres. 

We offer

• Medical insurance after the trial period ending;
• Team-building activities;
• 24 paid vacation days per year;
• 5 paid sick leave days per year.