Sriram Prashanth Email & Phone Number

Singapore

Singapore

Singapore

Liquidity Risk Reporting, Regulatory Reporting,Counterparty Risk,Statutory Reporting(DEC)

Singapore

• Developed a machine learning model aimed at detecting travel insurance claim fraud, significantly enhancing the monitoring capabilities of the claims team.• Conducted text analysis of claim descriptions with NLTK, distilling key insights that refined the fraud detection processes.• Utilized IBM Watson Studio’s generative AI capabilities for ML implementation.• Employed SMOTE (Synthetic Minority Over-sampling Technique) to address imbalanced datasets; achieved positive recall of… Show more • Developed a machine learning model aimed at detecting travel insurance claim fraud, significantly enhancing the monitoring capabilities of the claims team.• Conducted text analysis of claim descriptions with NLTK, distilling key insights that refined the fraud detection processes.• Utilized IBM Watson Studio’s generative AI capabilities for ML implementation.• Employed SMOTE (Synthetic Minority Over-sampling Technique) to address imbalanced datasets; achieved positive recall of >0.7• Included an additional anomaly detection algorithm to detect anomalous claims (IsolationForest)• Successfully identified that 1/3 of the suspicious claims churned out by the model were fraudulent or required loss adjustment, demonstrating the model's high accuracy and effectiveness.• Formulated a predictive model for estimating damage costs by employing techniques such as DBSCAN clustering and XGBoost, thereby assisting in better financial planning and resource allocation.• Utilized unsupervised learning models for damage estimations, considering various factors like accident location, collision type, and weather conditions, to optimize claims management.Web scraped insurance premium listings from competitor websites using Selenium to conduct a robust competitor analysis and maintain competitive pricing.• Applied Generalized Linear Models (GLM) to analyze and compare car premium calculations, leveraging distributions like Gamma, Inverse Gaussian, and Tweedie to derive relativity weightages.• Performed sample & KYC checks on suspicious claims• Tableau dashboarding and wrote python scripts to extract document information(PDF tables, websites) Show less

Singapore, Singapore

Was working in DBS-MOT-Risk Management Department(Middle Office Technology), Young Seeder Program• Performed data lineage analysis using data extraction tools (COLLIBRA, SAS)•Developed an automation script in Python for parsing large complex SQL queries (500 lines+ with nested expressions), extracting data lineage and automatically generating insert statements, ensuring accurate and efficient data storage and transfer.(src to target mapping)• Wrote a Python script to calculate… Show more Was working in DBS-MOT-Risk Management Department(Middle Office Technology), Young Seeder Program• Performed data lineage analysis using data extraction tools (COLLIBRA, SAS)•Developed an automation script in Python for parsing large complex SQL queries (500 lines+ with nested expressions), extracting data lineage and automatically generating insert statements, ensuring accurate and efficient data storage and transfer.(src to target mapping)• Wrote a Python script to calculate Somers' D statistic, contributing to improved rank correlation assessments and enhancing the decision-making related to predictive credit risk models.• Crafted custom text/excel processing scripts to enhance database management.• Delivered pythonic versions of SQL functions and refined SQL queries for optimal data mapping. Show less