Ms. Prashanthi Matam
Intelligent Fraud Detection in Credit Card Transactions Using AI
Abstract:
In an era of increasing digital transactions, credit card fraud poses a growing threat to financial institutions and consumers alike. This talk explores how Artificial Intelligence (AI) can be leveraged to intelligently detect and prevent both third-party and first-party credit card fraud with greater accuracy and speed. Using real-world datasets such as Stripe Radar's risk score predictions for payments and fraud probability data from Card Transaction Fraud records, we demonstrate how AI models can be trained to identify suspicious patterns, flag high-risk transactions, and adapt to evolving fraud tactics.
Our approach combines supervised learning with advanced feature engineering to enhance fraud detection systems. We will also highlight the unique challenges in distinguishing between third-party fraud (unauthorized use of a card) and first-party fraud (intentional chargebacks or misuse by the legitimate cardholder), and how AI can be tuned to address both. The session concludes with insights into model evaluation metrics, deployment considerations, and ethical concerns surrounding AI in financial security.
Profile:
a Senior software engineer with over 6 years of industry experience spanning AI/ML, full-stack development, and cloud infrastructure. I’ve built and scaled end-to-end data platforms and intelligent automation systems at companies like Discover and Goldman Sachs, and have hands-on expertise in tools like Python, Angular, AWS, and Kubernetes. My work includes integrating machine learning into production environments, enhancing user experiences, and contributing to dynamic, cross-functional teams.
I’m passionate about building impactful tech, mentoring early-career developers, and creating inclusive spaces where innovation can thrive. Whether it’s developing predictive analytics solutions or automating cloud-native systems, I love connecting the dots between code and real-world outcomes.