Juri Botnikov

Developed ML-models for internal customers: • CRM (propensity, churn), from classical to graph-based models • NLP for text classification, from agreeing business requirements to working prototype solution • NLP-based staff and service quality control for Call Center Supported different analytical projects, e.g. AutoML in secure environment

Developed ML-models for internal customers: • CRM models - new product targeting models based on internal transactional data and external data - premium products targeting models • Collection (model-based strategy optimization), from agreeing business requirements and raw data analysis to the set of models for different products and collection stages

Developed ML-models: • Antifraud graph-based models (for Retail Risk Department) • Models and algorithms for cash management & allocation for branches and ATM-network (for Cash Management Center) • Branch office staff evaluation models (for HR Department) Teaching mathematical statistics to Bank employees Managed Advanced Data Analytics team (up to 5 employees)

Developed retail risk models for Bank and its subsidiaries (incl. international branches): • app. scoring (incl. Credit Bureau and other external data). • anifraud • collection • CRM (response and income prediction for pre-approved loans) • AIRB models (PD/LGD/EAD) • retail credit portfolio models (NPL, provisions) Data- and model-driven optimization of lending process Created internal retail model development methodology Managed Retail Risk Modeling team (up to 8 employees)

Developed retail risk models for Bank and its subsidiaries (incl. international branches): • application scoring models (incl. Credit Bureau data) • retail antifraud models • collection models • data- and model-driven optimization of lending process • AIRB models (PD/LGD/EAD) • retail credit portfolio models (NPL, provisions)

Developed retail risk models: • application scoring models (incl. Credit Bureau data) • retail antifraud models • collection models • data- and model-driven optimization of lending process • retail credit portfolio models (NPL, provisions)

Developed retail risk models: • application scoring models (incl. Credit Bureau data) • retail antifraud models • data parsing and preparation

Developed predictive models for different external customers (finance, insurance, medicine, genetics, etc.) Proposing algorithms for internal software improvement (data preprocessing)

Probability Theory & Mathematical Statistics

Probability Theory & Mathematical Statistics