Skip to main content
Key Contributions
- Analyzed financial, operational, and environmental data for approximately 2,000 entities within CCB's framework of more than 30 indicators, assessing overall scores alongside minimum requirements for individual metrics.
- Used PCA to identify overlapping indicator information and construct compact representations for candidate comparison. Applied supervised learning to historically labeled cases, examining feature patterns associated with positive and negative assessments and adding empirical evidence to candidate assessment.
- Incorporated life cycle assessment principles into sustainability analysis, examining operating and environmental factors alongside financial performance.
- Tested the initial top 30% of ranked candidates against individual-metric and overall-score minimums, retaining approximately 28% of the full pool for further research and review.
- Reviewed selection and exclusion patterns across the pool, separating ranking cutoffs from threshold failures. Converted borderline cases, unmet requirements, and follow-up questions into a reusable checklist linked to the applicable screening rules.
- Delivered a methodology note explaining scoring logic, data sources, model behavior, and edge cases to product managers and risk officers assessing the method for platform use.