Description
The Internal Audit Department invites a specialist to join its team in the field of model and Data Science auditing. You will join a team of experts engaged in evaluating and improving machine learning models and data to ensure their compliance with high standards and business objectives.
Responsibilities
- conducting audits of statistical, ML models (including DL, LLM, neural network models, NLP) and automated algorithms (assessing the correctness of methodology, model development and application, calculations, and data quality)
- diagnosing issues in banking processes and automated solutions (models/algorithms), formulating recommendations for their elimination
- conducting validation tests, preparing reports and materials based on audit results
- developing audit approaches and tools (test automation, implementation of new methods)
Requirements
- experience in DS/validation/model risk/internal model audit (or related fields: analytics/model development with a focus on control and quality)
- proficient Python for data analysis: pandas, NumPy; visualization (matplotlib/seaborn) — at the level of diagnosing and explaining results
- mathematical statistics and ML fundamentals: correlation/statistical analysis, quality metrics, type I/II errors, understanding of linear models, trees/ensembles, basic neural network approaches
- practical experience in conducting validation/verification tests
- Apache Spark: proficient use of PySpark/Spark SQL for big data processing, job optimization (partitioning, join strategy, cache)
- skills in preparing clear reports for business/developers/validation teams (structure, conclusions, risk-oriented recommendations).
- skills in reproducible calculations (environment, requirements, fixing datasets/seed)
- instrumental proficiency in AI for analysis, generation, and automation
- analytical mindset, responsibility, communication skills, ability to work under tight deadlines and multitask
Will be an advantage:
- experience in developing/training ML/NN for applied tasks
- PyTorch or TensorFlow (for reading/checking implementation and training artifacts)
- experience in audit/internal control/model risk management (scoring/limits/anti-fraud/, PD/LGD/EAD, NLP)
- SQL (advanced), Git, Linux
- developing AI agents for applied tasks (including RAG)
Conditions
- employment according to the Labor Code of the Russian Federation
- guaranteed income plus bonus remuneration
- flexible mortgage loan discount equal to 1/3 of the Central Bank's key rate
- Voluntary Health Insurance from day one and preferential insurance for close relatives
- free SberPrime+ subscription, discounts on products from partner companies
- corporate pension program
- company-paid training: online courses in Sber's Virtual School, opportunity to obtain new qualifications
- employee referral program: you can invite familiar professionals to the team and receive a reward
- workplace: St. Petersburg, Shafirovsky Prospekt, 2
