Vakansiya haqqında
Əsas vəzifələr
Analyze existing data structure and assess data quality
Build data parsing and entity extraction pipelines to enrich customer profiles; apply normalization and masking in accordance with privacy regulations.
Building probabilistic models to extract demographic and behavioral attributes (supervised/weak supervision, calibration, confidence scoring) and writing the results to a feature store.
Identify data quality issues and propose improvement mechanisms. Perform feature engineering and model selection
Develop ML models for anomaly and pattern detection. Evaluate and optimize model performance
Building lookalike audience logic: embedding + nearest neighbors / gradient boosting / deep learning models to find similar users from seed audiences; sizing and updating segments
Tələblər
Bachelor’s degree in MIS, Informatics, Computer Science, Computer Engineering,Mathematics or related field
Strong proficiency in Python (Pandas, NumPy, Scikit-learn ,TensorFlow, PyTorch)
Minimum 3+ years of experience in Data Science or Machine Learning
Preferred: Adtech/Martech, CDP/CRM, lookalike/segmentation, recommender systems, fraud detection experience.
Python (Pandas, NumPy, Scikit-learn), at least one DL framework (TensorFlow or PyTorch).
SQL (query optimization), data warehouse understanding.
Big data tools: Spark/Databricks or equivalent (preferred).
Basics of model interpretability (SHAP/feature importance), calibration, bias/fairness.
APIs (REST/gRPC), containerization (Docker), orchestration (Airflow/Kubeflow or equivalent) – basic/intermediate level.English (advanced)
Strong analytical and problem-solving skills
Təsvir mənbədəki dildə saxlanılıb. Göstərilməyən şərtlər təxmin edilmir.
