initiated project management for my team. This time, being tight on budget, it was in Redmine. Tasks and documentation were organized & managed there. Though - despite customising efforts - not even closely as functional & efficient as Jira; - as ROC AUC, GINI and other data science precision metrics are good but do not always maximize ``bang for the buck'' - I wrote a ``cut-off'' optimisation tool in Python. It basically weighted, at a given cut-off, net expected revenues from GOODS to the cost of accepting BADS, given some details (LTV, utilization, funding, OpEx, LGD, ad-hoc parameters, etc.) and iterated till an overall revenue peak was found. It introduced a transparent automated way of setting efficient cut-offs. This one showed a potential for Approval Rate increase. But then ``corona-crises'' shortly lead to market rates' shifts and credit policy tightening. Focus shifted to Collection; - led various initiatives for external data utilisation: telecoms (2 sources), government (2 aggregate sources). Dealt with APIs, data format & structure, etc. Internal data sources were also pinpointed for `transaction' models' use. Various business requirements were written, several models built; - collaborated with Developers on Decision Engine (back) and Application Forms (front) tuning (BRDs, FSDs, change requests...), with Business units - on business- process improvement (UML process «flow-charts»); - introduced a loan portfolio valuation model - in order to valuate another creditor's portfolio that was on sale. An `income based' approach was utilized; - there were other numerous operational & analytical tasks, mostly on SAP CRM.

Moscow Business Analysis, Marketplace. 10 people in project (DB, back & front devs, testers, designers). PWA application (React + Redux, Spring Boot & Java 9, Postgresql, Intellij IDEA, AWS). ------- Key responsibilities: - project roadmap setting (Confluence), performance metrics setting & tracking, release controls, competitor's analysis, ect.; - product scope (Jira): backlog filling with `user stories', BRDs/FSDs, rarely - transforming BRDs to tech specs for Devs, features prioritising (RICE), Lean; - leading weekly Scrum meetings; - collaborating with Designers (outsource); - created several detailed flows (wireframes in Sketch+InVision, Figma), became quite savvy in UX (Apple's HIG, Google's `Material'); - matched lenders' Decision Engine's rules to Finport's APIs and internal workflows; - collaborated with Credit Bureaus, setting APIs; - created a Data Model concept (CRM-like, including log data, external data feeds), basic datamarts for reporting; - devised a Knowledge Base in Confluence; - initiated a digital marketing plan (SEO), user research. Key achievements: - product strategy, full backlog, `growth points' identification; - setting clear & realistic tasks, estimated and, if needed - functionally decomposed; - setting `A/B testing' process management culture with a feedback loop; - launch of MVP (late `alpha').

RSB24 (ex-ContactBank), ------- The main idea was to try to reinvent bank's sustainable position by installing points of growth with controllable levels of risk. Key responsibilities: - harmonizing bank's risk profile across multiple risk-taking businesses, working closely with business on changing product pipelines with respect to the risk strategy; - automatization of risk processes (risk team itself or with IT team); - SAS installation project - BRDs for data inputs, Decision Engine; - DWH installation project - BRDs for Data Lake, MDM, Risk datamarts; - building scorecards for SMEs (setting workflow with decisions given within 5hrs); - looking after retail clients and up-selling them products; - collaborating with 3 major Credit Bureaus, `Big Data' providers (mobile operators, transaction servicing companies, social networks' data aggregators etc.) in order to improve credit risk models' precision and stability; - reconfiguring `antifraud' procedures; - collaborating with Collections (in-house and third party) in order to improve recoveries (collection scorecards), consequently retuning LGDs; - analysing economic/industry/regulatory developments, establishing relationships that could explain portfolio risk & macro-economic correlations that drive net credit losses, liquidity, interest rate and FX risk; - member of ALCO, Credit, Change and Risk Committees, as well as of `horizontal' projects, reporting to senior management, accountable to the Supervisory Board. Key achievements: - building a team of risk professionals with different backgrounds and experiences; - all business projects were provided with respective risk parameters/models;

Responsible for managing bank's risk profile set by the Supervisory Board and acceptable to the regulators. That includes quantitative and qualitative assessment of all types of risks: credit (retail, corporate & banks), market (interest rate, FX), operational and liquidity risk, as well as engagement with key personnel in order to execute efficiently Bank's strategies. Key responsibilities: - participating in new products' reviews, ensuring compliance with current policies & procedures, establishing & reviewing basic underwriting standards (decision trees); - controlling loan portfolio quality - tuning approval processes (SQL procedures, credit scoring models in R, underwriting workflows), setting performance metrics & triggers, identifying and analysing trends and concentrations, evaluating risk performance under various stress conditions, reformulating workflows (UML); - working closely with Collections to ensure the efficacy of the Bank's end-to-end controls for all aspects of customer behaviour (incl. collection scorecards); - participating in budget setting, ALM, monitoring `macro' developments that drive risks, scenario analysis, ``connecting dots'' (short & medium term forecasts); - DWH installation project - BRDs for Data Lake, MDM, Risk datamarts; - leading a staff of 50+ (optimized to 30+), 5 functional directions: ALM, risk analytics & reporting, credit risk verification, risk methodology, operational risk; - business oriented KPIs - maximising profits, keeping risks at acceptable threshold; - member of ALCO, credit committee and `horizontal' projects (FTP, portfolio approach for retail loans provisioning, international credit rating, ect.); - providing advisory support & reporting to the Supervisory Board, CEO & regulator. Key achievements: - comprehensive risk-management system, integrated into all business lines/levels; - team of inquisitive and goal-oriented risk professionals (KPIs matter!!!); - working under highly risk-tolerant bank's strategy, still outperforming (given portfolio risk metrics) bank's peers that were selling by far less risky products; - receiving the highest possible (for local market) bank's credit rating (Moody's). Key failures: - inability to influence key business points of Bank's long-term strategic plans; - underestimating ``human factor'' on top executive & shareholder level.

Quantitative loan portfolio management, loan transaction approval methodologies: - tuning and fine-tuning (optimization) underwriting settings (Fair Isaaс's Blaze) according to communicated strategy (approval rate boosting, default level decreasing, RAROI): functions, policies, scorecards, cut-offs, verification workflows, limits (client/transaction), credit bureau ect. I.e. introduction in 2009 of a new scorecard and cut-off let the bank double automatic approvals while keeping the default rate stable. Before deploying new strategies to full production, `offline' simulations are run on clients' DB (also segmented by products, vintages, ect.). `Focus groups' (A/B testing) are prepared to check the efficiency of current settings; - scorecards building (R, newbie in SAS): monitoring current scorecard's efficiency (ROC, LIFT, GINI...), building & validating new ones (algorithms: log- regression, NN, GAM...), choosing most appropriate (accurate & stable), associating a cut-off, running Champion/Challenger. Currently application and behavioural scorecards are used (too little data for Collection SC); - writing BRDs/FSDs for Devs (in English) mostly on Decision Engine procedures & data manipulation, assisting at releases for the risk part of the SW. Participating in Marketing & Finance: - precising & segmenting target population (cross&up-selling), risk-based pricing, forecasting, optimization, stress testing ect.; - monthly provisioning for Non Performing Loans with statistical models (Roll- Rate, Markov, LGD with respect to recovery) and according to IFRS 39; - budgeting, forecasts and 'triggers': allowances for impairment losses, response rate for x-sell, OpEx, risk indicators' triggers setting (FPD, 30+, 90+...). Reporting & operational stuff: - ad hoc and regular (d/w/m/q/y) reports (SQL+Excel+Powerpoint) to local senior management (Supervisory Board), other bank units, auditors and regulatory bodies, Group's HQ. Reporting covers approval process (underwriting), Fraud, Provisions, Collection, Macro environment (GDP, Unemployment, FX ...) ect.; - participating at HQ conferences, regular call-meetings (in English), personally reporting to Group's CEO in 2010, received BACEE certificate; - creating documentation on risk processes (UML), managing a staff of 4.

Bank VTB (Belarus), Minsk (VTB Group) ------- Loan portfolio monitoring, FX & interest rate risk measuring and stress-testing (Excel+VBA). Hold `implementation act' for Historic & Monte-Carlo Value at Risk simulation program. Later, National Bank's disposition on market risks was `inspired' by the internal document regulating described methodology.

Lectures and seminars to Master degree students (Finance & Banking), author of the course `Credit Risk Management In Banking'.

Banking software analyst (Treasury Management, outsource project).

Bachelors Degree

Master of International Economics