Panaja Kasarla

I'm currently building fraud analytics and risk operations strategies at AT&T, using complex SQL queries and Python scripts to uncover transactional anomalies and strengthen payment fraud controls.

My work has improved fraud detection accuracy by 20%, reduced false positives by 15%, and increased alert precision by 10% through feature development, experimentation, model evaluation, threshold setting, and statistical analysis.

Previously at Innominds, I queried more than one million payment records, automated Python data processing, and reduced SQL query runtime by 25%. I also evaluated detection rules, supported A/B experimentation, and developed fraud analytics frameworks for payment fraud and account takeover detection.

I translate fraud typologies—including synthetic identity, business email compromise, stolen payment instruments, and first-party misuse—into documented rule logic, operational controls, and practical fraud prevention plans.