Description
The Data Research Department of the Compliance Directorate is seeking a Head of Data Research.
The primary task will be developing ML models for the process of identifying and investigating cases of money laundering. Various model execution platforms are used – online (at the moment of transaction) and batch calculation across the entire client base. We use ensembles of gradient boosting models, neural network architectures on sequences of events, time series, and graph data structures.
Responsibilities
- Decompose tasks and manage their timelines;
- Collect samples for training and validating models;
- Test various models for solving the task (classification, regression, anomaly detection), select the best model;
- Formulate requirements for data marts for the engineering team;
- Conduct code reviews;
- Prepare presentations to defend results;
- Analyze the results of monitoring model performance in the staging environment;
- Participate in candidate interviews.
Requirements
- Relevant higher education;
- Good knowledge of probability theory and statistics;
- Knowledge of Python and core libraries (numpy, pandas, scipy, sklearn);
- Confident command of popular machine learning algorithms and gradient boosting libraries (LightGBM, CatBoost, etc.);
- Experience in developing and deploying neural networks on PyTorch.
- Ability to process large volumes of data with PySpark;
- Cases of applying neural network models in NLP, Time Series, Graph tasks;
- Experience in mentoring or leading a team.
Conditions
- Free fitness centers in the office;
- Private health insurance from day one and preferential insurance for close relatives;
- Favorable mortgage for every employee and preferential loan terms;
- Discounts on products from partner companies;
- Online courses in Sber's Virtual School and training at the Corporate University;
- DS&AI community - regular exchange of knowledge, experience, and best practices, interactive lectures and master classes from leading universities and experts of tech companies.
