Antony Dhanapaul

Data Scientist with 5+ years of experience delivering AI/ML solutions across financial services, insurance, and telecom industries. Skilled in developing real-time anomaly detection, AML monitoring, and AI-driven risk scoring models to enhance fraud prevention and regulatory compliance. Experienced in designing and building scalable data pipelines, ETL workflows, and production-grade ML systems using Python, PySpark, SQL, and cloud platforms (AWS, Azure, GCP).

Adept at translating complex business problems into data-driven insights that improve decision-making, operational efficiency, and model performance.

Data Scientist - Sanizer, Texas, USA - Sanizer, Texas, USA

(2024-02)

• Led end-to-end design and deployment of AI/ML fraud detection and AML monitoring models (logistic regression, random forest, gradient boosting, anomaly detection), reducing fraudulent transaction leakage by 22% across high-volume payment flows and aligning with enterprise data science best practices.
• Built and operationalized real-time anomaly detection models aligned with ISO 20022 and RTP frameworks, improving suspicious activity detection accuracy by 18% and ensuring full BSA/AML and FCC compliance through advanced data analytics and model governance.
• Reduced false positives by 35% through fine-tuning Actimize scenarios, thresholds, and machine learning parameters, maintaining >92% fraud detection accuracy and accelerating investigation turnaround time.
• Automated model monitoring and tuning workflows using Python, SQL, and Splunk, cutting manual review efforts by 25+ hours/month and enabling continuous optimization of detection and compliance strategies through automated data pipelines.
• Developed ALM/stress risk scoring, sanctions screening, and KYC remediation frameworks, automating decision-making, reducing analysis workload by 45%, and achieving a 30% faster case closure rate with improved global FCC regulatory reporting accuracy leveraging machine learning models.

Data Scientist - Crusade technologies, Texas, USA - Crusade technologies, Texas, USA

(2020-01 - 2022-04)

• Designed and deployed a Roadside Assistance Digital Transformation system, automating liability assessment and FNOL using predictive data modeling, cutting claim closure time from 1–5 weeks to 2–3 days and saving $7–43M annually.
• Built ML-based claim fraud detection models using supervised learning, enabling automated fraud scoring and early intervention; reduced fraudulent payouts by 20% within a year and improved compliance with audit requirements.
• Implemented ML-driven churn prediction models, enabling data-backed retention campaigns that improved policy renewal rates by 8% and increased overall customer lifetime value through predictive analytics.
• Optimized premium pricing and underwriting models using regression and survival analysis, improving pricing accuracy by 12%, enhancing risk-based pricing, and driving healthier loss ratios.

ML Engineer - Cognizant, Hyderabad, India - Hyderabad, India

(2018-02 - 2019-12)

• Built and deployed an automated RFP proposal application, cutting 15 manual hours per proposal and saving 20–30 hours weekly across a sales team of 10+, improving proposal turnaround time by 40% through data automation and process optimization.
• Designed ML-driven customer segmentation and scoring models, boosting proposal win rates by 18% and driving $32.0M in incremental annual revenue through data-driven insights.
• Built predictive data enrichment models (financial health, purchase history, credit risk) that improved customer data accuracy by 25%, reducing rework and accelerating proposal approvals using statistical modeling.
• Automated ETL workflows and feature engineering pipelines, reducing data preparation time by 80% (days → hours) and enabling 30% faster response to RFP/RFA opportunities through data integration practices.

Masters - Information System - Wilmington University, New Castle, DE

Bachelors - Information Technology - Anna University, Chennai, India