Description
Our team manages model risk arising from incorrect model operations. This risk can lead to financial and reputational losses.
Primarily, we work with models on tabular data. We also analyze NLP models. We write risk assessment methodologies, develop validation tests in Python. During validation, we delve into the business process and build alternative solutions. We also set up and conduct regular model monitoring.
We automate processes, both with the help of agents and by using non-agent technologies and internal platforms.
Responsibilities
- assessing the quality of Classic ML models: uplift, response, multiclass, multilabel, as well as assessing model performance and its applicability in the business process, verifying the model's effect on the business
- researching modeling and validation approaches, improving methodology, applying advanced technologies, and disseminating best practices
- automating processes (validation, monitoring, etc.), both using agents and without.
Requirements
- higher education in technical / physical-mathematical sciences
- knowledge of classical machine learning
- work experience and understanding of approaches used in campaigning: uplift, PyBoost, response models, propensity models
- understanding the basics of A/B testing.
- knowledge of Python and data analysis libraries: PyTorch, HuggingFace, and Transformers
- knowledge of PySpark, SQL
- experience in model analytics and managing the model stack in a business process
- knowledge of the basics of NLP.
Will be an advantage:
- experience in building agents/agent systems, understanding their principles of operation, communication, and orchestration.
Conditions
- comfortable modern office: Moscow, Kutuzovsky Prospekt metro station
- work format– office
- annual salary review, annual bonus
- corporate gym and recreation areas
- more than 400 educational programs from SberUniversity for professional and career development
- onboarding program and manager assistance at the start
- extended VHI (voluntary health insurance), preferential insurance for family
- mortgage for employees
- free subscription to SberPrime+, discounts on products of partner companies
- referral bonus for recommending friends to Sber's team.
