nove8 9 дней назад
Data Scientist
- Удалённо
- Опыт от 2 лет
- Полная занятость
PythonSQLPostgreSQLData ScienceData AnalysispandasretentionNumPyLTVBigQueryMarketing AnalyticsScikit-learnstatsmodelsPredictive Modeling
Описание
Responsibilities
- Developing, validating, improving, and supporting key marketing forecasting models (long-term LTV, revenue, retention).
- Building and maintaining reliable time-based validation and backtesting approaches.
- Working with incomplete and censored cohort data, considering their maturity.
- Analyzing transactional and event-level data, checking data semantics, and identifying quality issues.
- Monitoring model performance, diagnosing prediction errors, and continuously improving quality.
- Building and supporting datasets and features, supporting A/B tests and statistical interpretation of results.
- Moving models from research to regular production calculations and gradually expanding modeling to other areas, including risk models.
Requirements
- 2+ years of experience as a Data Scientist or Data Analyst, or proven experience in implementing end-to-end analytical or predictive projects.
- Strong knowledge of SQL, including experience with Google BigQuery and/or PostgreSQL.
- Proficient Python for analysis and modeling with libraries pandas, NumPy, SciPy, scikit-learn, statsmodels.
- Deep understanding of mathematics, probability, statistics, statistical inference, and hypothesis testing.
- Understanding of classic statistical and ML approaches and experience in building predictive models.
- Confident model validation skills (time-based validation, backtesting, data leakage, overfitting, model stability).
- Understanding of censored and incomplete data, cohort maturity, and basic survival analysis concepts.
- Understanding of subscription economics and key marketing metrics (LTV, CAC, ROI/ROMI, payback, retention, conversion rate).
- Experience with large transactional and event-level datasets, analyzing discrepancies and biases in data, translating analytical results into business insights.
Nice to have
- Experience with subscription products, recurring billing, and payment data.
- Experience with long-term forecasting, survival/hazard models, and other time-dependent approaches.
- Experience with Stripe, PayPal, or similar payment providers.
- Familiarity with dbt/Dataform, Airflow, Tableau, Amplitude, AppsFlyer, Meta Ads, or Google Ads.
- Knowledge of production ML processes, model monitoring and retraining, and classic ML models.
We offer
- Unlimited days off outside the office.
- Compensations for wellness, psychologist, language courses, and coworking.
- Medical insurance.
- 100% sick leave payment.
- Team buildings and offsites (Spain, Italy, Poland, Portugal, and others).
- Merchandise and gifts for holidays.