Sachin V S Email & Phone Number

Model Risk and Risk Analytics; Mortgage Insurance

Kingston, Ontario, Canada

Intelligent financial statement fraud detection via NLP

CISC 251 - Data Analytics

• Member of the Global Data Science team involved in creating automated modeling components in D&B - Analytics Studio, across multiple credit risk domains.• Led the development of an end-to-end insurance analytics package for identifying risky and targetable customers; from gathering requirements to creating interactive dashboards. Focused on loss-ratio modeling, and utilized techniques such as quantile binning, variable clustering, and gradient boosting to highlight specific D&B data categories as key predictors. Applied Boruta feature selection to achieve a significant reduction in MSE by 7%.

Noida Area, India

• Helped rewrite legacy APIs for an InsurTech client (Property and Casualty) using different tools and services in a Microsoft tech stack.

Chennai

• Utilized Apache Tika to mine text content from web-scraped company annual reports and built predictive classification models (via TF-IDF and other methods) of future monetary performance; investigated 3 financial ratios to establish ground truth.• Explored the impact of Altman Z-score (financial distress) to detect false financial statements using EMIS private company data. Applied feature combination and threshold tuning to improve F1 score by 5 percentage points.• Presented procedures and results regularly to senior managers and partners for feedback.

Master Of Science (M.Sc. By Research), Computer Science, 4.3/4.3

* Grad student in the Queen's University School of Computing, under the supervision of Dr. David Skillicorn. * Thesis: Revisiting Fraud.

Bachelor Of Technology - Btech, Computer Science And Engineering, First Class

Activities and Societies: IEEE and regular participation in numerous departmental initiatives.