Загрузи резюме, чтобы видеть мэтчи с вакансиейЗагрузи резюме, чтобы видеть мэтч с вакой
подготовься к отклику
ai-инструменты
Чтобы адаптировать резюме под вакансию или составить сопроводительное письмо, загрузи резюме
О рекламодателе
ОБЩЕСТВО С ОГРАНИЧЕННОЙ ОТВЕТСТВЕННОСТЬЮ "ЦЕНТР НАЦИОНАЛЬНЫХ ИНТЕЛЛЕКТУАЛЬНЫХ СИСТЕМ" ИНН: 9704271170
описание
We are seeking a Senior Data Scientist for a leading AdTech company whose sell-side omnichannel platform uses AI and 100+ billion daily first-party data signals to deliver hundreds of millions of ads every day. In this senior, hands-on role, you will work at the core of a high-volume, real-time programmatic system processing over 2 billion requests per day. Your work will directly drive campaign optimization at scale through production ML systems, not offline research.
задачи
Analyze large, complex datasets to uncover hidden patterns and opportunities
Build and own end-to-end, highly scalable data pipelines
Develop state-of-the-art ML models and algorithms for monetization
Provision data models for real-time (online) processing
Shape new data products and continuously improve existing ones
Partner with Product, BI, Analytics, DevOps, R&D and Marketing to deliver end-to-end solutions
Own full releases to production, from requirements analysis and testing through monitoring, measuring results, and responding quickly to critical issues
требования
B.Sc. or M.Sc. in Computer Science, Engineering, Mathematics, Statistics or a related field
6+ years as a Data Scientist with a strong grounding in ML concepts and models
Hands-on Python experience with data science and ML libraries (e.g., XGBoost, LightGBM, regression and classification, recommendation systems)
Solid SQL and database knowledge for data retrieval
Experience with monitoring and alerting tools (e.g., Grafana, Kibana)
Будет плюсом: Knowledge of the programmatic advertising landscape (DSPs, SSPs, exchanges), which is a significant advantage; Experience with relational (e.g., Vertica, VoltDB) and non-relational (e.g., MongoDB) databases; Experience with PMML