Anton Lebedev Email & Phone Number

Москва, Москва, Россия

Mar 2021 – present: Data Science Director – Services (Infrastructure) Unit• Implementing AI Transformation to business processes of the Unit: cash flow collection, estate and exploitation, construction, supply, document management, security, cybersecurity • Creating and growing products with AI: Auto scheduling of cash flow collectors, Energy management and electricity consumption optimization, Automatic generation of construction documentation for offices reformation, Auto replenishment of shipment at the Bank objects, Criminogenic profile of location of automatic teller machine (ATM), Digital shape of DDOS attack• Managing Data Science team, 11 DS + 2 MLE• Understanding business needs, estimation of possibility of AI solution development and implementation, tasks definition for data scientists, setting deadlines and quality control• Coordination Agile team activities – linking business owners, analysts, developers, engineers• Managing and controlling main KPIs of AI Transformation of the Bank: AI Maturity Index• Making presentations for b-1 level

Москва, Россия

Jun 2019 – Feb 2021: Data Science Director – Sales Net (Branches) Unit• Implementing Sberbank AI Strategy in business processes of the unit• Managing and controlling main KPIs of AI Transformation of the Bank: AI Maturity Index• Designing and developing AI Solutions for current business clients• Organizing A/B testing and estimating financial results of implemented AI solutions• Leading three teams of data scientists who work for three business departments: direct sales, payroll projects and technology of sales managers• Understanding business needs, estimation of possibility of AI solution development and implementation, tasks definition for data scientists, setting deadlines and quality control• Coaching and skills development of team members• Making presentations for b-1 level

Moscow, Russian Federation

• Communicated with internal business partners to collect and clarify the business goals and transform them to programming language, taken decisions about possibilities of statistical modeling• Performed full cycle model design and development which includes but is not limited to identifying objectives, sampling, testing/validation, calibration, and monitoring performance• Selected and prepared specially formatted data sets (TARADATA, SAS) for predictive modeling• Tested new information about Bank clients in order to improve current models• Conducted technical and quantitative researches for predictive modeling purposes such as outlier analyses, stability analyses, etc.• Built custom, explainable statistical models – logistic regression and decision trees to find clients preferences and ranking them, also built models for non-retail business segment (HR)• Monitored and maintained implemented models, automated scoring process• Set scheduled execution of model scoring scripts, adjusted automatic reporting process• Led an ad-hoc analysis of business problems of various scope and complexity• Developed additional data marts (Hadoop) for raw data sources such as social networks data and web data, promoted data among different software (SAS, TERADATA, Hadoop)• Worked with other data engineers and modelers on quality assurance of source data and systems• Shared knowledge among colleagues and tutored new ones in order to help them to work in a team• Done information security administrator duties • Apply python 3 for text mining task, social network text data analysis (in progress)

Moscow, Russian Federation

• Developed and supported data mart for SAS Marketing Automation• Created and maintained ETL processing• Analyzed business tasks concerning growth of CRM data mart, made suggestions about creation of new entities and their content• Controlled scheduling process, deployed jobs, traces correctness of scheduled ETLs and monitored performance• Optimized the resource usage and performance of DI jobs, realized an incremental load techniques, initiated an initial load and planned table population, utilizing advanced loading technologies and methods• Prepared implementation packages for production environment with source codes, metadata and implementation instructions• Maintained and actualized data mart documentation with ETL algorithms description and table descriptions• Settled data validation issues, fulfilled user’s queries and incidents• Improved current versions of information maps• Connected with CRM managers, marketing company’s analysts to discuss data mart content and issues of users satisfaction with the data mart; considered performance bottlenecks• Cooperated with DWH architects and other data mart developers in order to settle issues of source availability and performance tasks

Moscow, Russian Federation

• Designed various monitoring and inspecting reports of macroeconomic performance• Made automation of new and existing reports• Created user-friendly stored processes for management to provide a convenient usage of reporting facility via web applications, such as SAS Web Report Studio• Applied SAS ODS and/or Microsoft Office (Excel, Power Point) to output the results• Improved the efficiency and correctness of current automation processes using advanced methods of SAS programming• Validated the quality and accuracy of receiving data• Developed several data marts of financial reporting data and financial activity monitoring data • Created and maintained ETL processing• Deployed jobs, scheduled work-flows and maintained the user-defined logging• Coordinated the data mart development process, approved attributes and a table structure, regulated the scheduling process and was responsible for data updating • Collaborated with different managers to clarify business tasks and discussed ways of effective solutions• Conducted an ad-hoc analysis

Moscow, Russian Federation

• Supported and revised existing reporting automation process• Invented and enforced automated reporting data validation modules, including graphical panels of detected data inconsistencies and irregularities• Executed processes of creation of data samples • Explored different data sources and searched outliers and other peculiarities• Maintained existing SAS-DWH of business performance reporting• Analyzed performance indicators such as Approval Rate, Loan Portfolio and historical indicators such as Roll Rates, Vintage analysis• Cooperated with data providers to quickly reply on various data issues and discuss the ways of improving data quality and integrity• Developed score cards for several types of credits, facilitated a GINI analysis and a ROC-curve analysis• Performed ad-hoc queries and tasks