Asst Vice President - Data Science
- Spearheaded efforts in building XGBoost based propensity model to identify emerging companies based on the company’s industry, growth, founder’s profile, active investor’s profile, and investment activity, enabling early-stage banking relationships and long-term partnerships which has resulted in $20M in first-year incremental revenue
- Achieved successful development of a sophisticated look-alike clustering model with FAISS algorithm that utilised a semantic search mechanism, leveraging filmographies , financial data, and risk information from external dataset (Dun & Bradstreet) to effectively identify new-to-bank companies with striking similarities to our existing banked customers. Notably, this implementation resulted in a remarkable $5M in annual incremental revenue in 12 ASP Markets.
- Designed and delivered a location-based data product, empowering relationship managers with necessary information from external data sources (D&B, Refinitiv, Factset) to identify new business prospects within close proximity of their existing client base, speed up and solidify the business development and customer service activity which is projected to save 70% in the frontlines client development activity.
- Led the development and implementation of and ML-driven risk indicators prioritization framework for empty shell companies identification across 12 Asian markets by analyzing payment patterns, customer demographics, and RM feedback notes, projected to 60% reduction