Sayantan Ghosh

Establishing & leading a team of risk professionals to develop profitable policies spanning underwriting, limit assignment, increasing share of repeat customers, etc. Liaison with other business facing teams with focus on how Data Analytics can produce insights to drive tangible business outcomes.

Leading a team of risk professionals to develop profitable new account underwriting strategies for international markets. Collaborate with multiple partners including digital analytics, modeling, operations & technology teams to implement, monitor and manage strategies & processes.

Leading a team of risk professionals for monitoring, developing and implementing various Responsible Lending Actions for the US Consumer Portfolio - Responsible Lending Actions include line reductions, proactive cancellations, removing the revolve feature or placing a spending limit on a charge card product leveraging internal and external risk scorecards. Leading extensive analysis into card member behavior to optimize strategies & capabilities to ensure end-to-end management of portfolio credit risk. Partnering with multiple global partners across International, Bank & Regulatory bodies and Operations to drive implementation of multiple global initiatives.

End to end ownership of Data Matching Algorithm (inclusive of data standardization, optimal candidate selection & fuzzy scoring techniques) – utilized across various enterprise wide verticals such as New Accounts, ID Generation, etc. Intelligent Customer Contact - solving for features such as the Best Phone Number for a Customer through effective use of Machine Learning capabilities such as XgBoost. Involved in enterprise wide leadership initiatives such as hiring of new talent.

Mining for the optimal & most granular industry for a merchant using ensemble frameworks such as Random Forest & Pseudo Relevance Feedback on Bing Search Snippet Data.

Marketing Segment Optimization & Merchant Prioritization – Strategic analytics for enabling offers to cater to the right set of customers through the right set of merchants.

Identification of merchants most likely to suppress American Express plastics through effective utilization of various modelling & analytical procedures (like K-Nearest Neighbor, ARIMA, Gradient Boosted Decision Trees, Z-Scores, etc.).

Amalgamation of third-party SKU level information with proprietary closed loop data features to come up with relevant product recommendations to customers of a retail chain through Collaborative Filtering.

Logistic based spike analysis and forecasting future trends to identify problematic servers.