1. Customer data analysis, risk policy and anti-fraud strategy - consumer lending- Managed customer acquisition and risk strategy for new API channels; Negotiated with suppliers and deployed pre-screening rules; Portfolio management, cohort analysis, risk metrics monitoring and strategy fine-tuning- Developed anti-fraud rules and assessed impact of approval rate; Investigated the abnormal data of face recognition score and fixed the data; Enhanced 3rd party data usage to improve fraud management capability- Formulated risk policy of telemarketing channel; Standardized the system flow of credit application and whitelist verification; Reengineered the process of system interface invocation, data exchange and data storage- Assessed impact of the global risk policy changes on each channel (customer group); Consolidated and analyzed the data to demonstrate the justification of getting exemption for specific channel (customer group)- Optimized cash withdrawal rules; Performed reject reason analysis and investigation to find the root cause of the same credit customer group having different approval rates in different withdrawal channels; Studied the reason and potential risk behind customers’ behavior of withdrawing cash in different channels- Designed system interface to the banks, coordinated the risk rejection rules with the banks; Analyzed bank side rejection reasons, optimized the capital routing rules- Explored and developed new risk decision engine variables; Adjusted existing variables’ definition and simulated the result of rules affected;- Conducted risk decision engine variables’ accuracy verification and stability monitoring; Investigated problematic variables and designed repairing solution by collaborating with IT covering requirements, UAT and production- Fraud case investigation (impersonation fraud, intermediary application, telecom fraud); Complex network analysis; Operational risk identification, access control and authority management

1. Customer data analysis, risk rating and traffic allocation - online loan application- Built up customer risk segmentation model and traffic distribution strategy; Evaluated distribution effectiveness and analyzed customer conversion rate; Produced waterfall reports for customer screening- Designed and optimized the whole credit application and decision process, including auto-approval by decision engine, manual approval, overdue customers re-apply approval rules, blacklist mechanism etc.- Reviewed and realigned risk policy, tracked result of policy deployed; Designed and developed risk variables- Developed credit risk model, collection model, anti-fraud rules and applied into the online lending business- Enhanced overdue debts collection strategy, including internal collection arrangement and result tracking, outsource collection incentive scheme design and leads allocation management- Designed and maintained risk data platform, managing customer application info, device info, event tracking data and third-party data; Designed data collecting and updating mechanism- Worked with large datasets: data collection (obtaining data from a variety of sources e.g. databases, flat files, data interface), data cleaning, data validation, data exploration, data standardization, feature engineering

1. Modeling and risk policy - Credit card- Credit risk model development: model parameters definition (GBIE), roll rate analysis, vintage analysis, observe window and performance window, sample data selection- Model derived variables construction, variable binning (fine binning/coarse binning), WOE transformation, variable selection, model development and validation, model implementation and stability monitoring- Social network analysis(anti-fraud): nodes selection, exceptional data cleaning, fuzzy matching, network generation and partitioning, network triggered risk rules and scoring- Discussed and established social network anti-fraud strategy; Developed individual applicant anti-fraud model- Explored credit card fraud “cash-out” transaction data and defined fraud “cash-out” customer profile; Produced “cash-out” customers’ segmentation and follow-up strategy- Created and verified risk modeling data mart for AI/ML; Developed PBOC (credit bureau) report data mart2. Marketing strategy - Credit card and personal installment loan- Drew up the risk criteria and targeting strategy for transaction installment loan product; Developed propensity model for transaction installment loan- Worked out the telemarketing execution plan based on the volume of customers’ large amount transaction data and the manpower situation of the outbound call team; Included also the SMS marketing program- Designed and maintained card campaigns process of leads generation, campaign evaluation and risk monitoring- Developed credit card activation propensity model, activation outbound call strategy and A/B testing approach3. Data warehouse and data mart- Data warehouse upgrade project design and test; Data governance, data quality and data platform improvement- Designed risk and marketing data mart with relevant ETL processes for MIS reporting and analytics purpose

1. Risk policy and data analysis - consumer loan, mortgage, small and micro enterprise owners’ loan- Conducted risk criteria data analysis (online loan), risk policy development and decision engine implementation- Implemented post-loan risk warning strategy, consolidating internal personal loans overdue records and risk warning data pushed from PBOC (credit bureau); Monthly report generation and reporting process automation- Overdue collection strategy analysis and reporting; Outsource collection leads allocation planning, result tracking and incentive scheme adjustment- Maintained dashboard reports of flow rate analysis, vintage analysis and accumulated bad debts estimation, etc.- Loan suspicious fraudulent transaction data detection, investigation and fraud prevention (suspend merchants); Reviewed and optimized the payment detection logic and algorithms, balancing alert volume and risk coverage2. Credit bureau and regulatory reporting- Managed and streamlined the PBOC (credit bureau) reporting process and system; Fixed exceptional report data- Ensured that regulation reporting is in-compliance with guidelines; Attended to other regulators’ ad hoc requests

1. Marketing propensity model and strategy- Marketing propensity model development and validation for personal loan, foreign exchange products, etc.- Developed “Next Best Product” strategy to maximize the sales / revenue / profit- Implemented products display strategy for new internet banking registration customers in different stages2. Campaign effectiveness evaluation- Calculated uplift using TEST and CONTROL groups; Evaluated campaign effectiveness, including offer effectiveness, model effectiveness, channel effectiveness and channel synergy (DM / EDM / TM / SMS)- Marketing campaign result tracking of response, sales and revenue for different customer segments3. Retail customer insights and BI reports- Leveraged data analytics to perform market opportunity sizing, provided insights and campaign ideas for marketing strategy, e.g. to grow the personal internet banking registration and activation- Data-driven deep dive analysis on customer behavior for marketing and customer relationship management, e.g. studied the reason of “foreign exchange product purchasing drops in internet banking" using event tracking data4. Campaign data management & CRM- Led a team of 7 staffs to execute credit card campaigns: targeted leads generation, fulfillment, tracking reports; The campaigns included acquisition, activation, targeted spending, bonus point rewards, annual fee waiver, etc.- Implemented inbound call center product sales strategy; Fine-turned retention offers to reduce customer churn- Liaised with marketing and risk departments to draft the retention strategy, exclusion / inclusion criteria, segmentation and product assignment matrix (repricing / limit increase / bonus points / annual fee waiver, etc.)- Improved computing performance of data process by program code optimization and process reengineering5. Data warehouse and data mart- Data warehouse and data mart development, including data extraction, transformation, standardization

1. Insurance actuarial valuation data mart and BI reporting- The process was running on mainframe, providing data for reserve calculation of life insurance policies- Maintained the actuarial valuation data mart, with related data processing programs- Captured the insurance policies movement; Developed BI reports of premium analysis report, new business analysis, accident claim analysis and continuous mortality investigation

1. Bank work flow image system- This system was an OA system based on J2EE architecture with DB2 database. We developed several projects for different business processes, including credit card application, call center, mortgage loan, etc.- Business analysis, database design, coding, unit test, SIT (system integration test), UAT, documentation2. Global payment system- The original system architecture was provided by vendor. We worked on system customization (C++), integration, day-end batch data processing and dataset transfer between host system and global payment system