Rajat Agarwal

Papers

Total Citations

H-Index

About

Rajat Agarwal is a leading researcher at the intersection of machine learning, cybersecurity, and online advertising integrity. His primary research focus is the detection and mitigation of fraudulent, non-human traffic in digital advertising ecosystems—a critical challenge affecting billions of dollars in ad spend annually. Agarwal’s most notable contribution is the development of SLIDR (SLIce-Level Detection of Robots), a real-time deep neural network model described in his 2023 paper. This system uses weak supervision to train on large-scale, noisy data, enabling it to detect robotic traffic with high precision while adapting rapidly to evolving adversarial patterns. By operating at the slice level rather than the user level, SLIDR achieves both scalability and granularity, allowing advertisers to filter out invalid traffic in milliseconds. Though early in its citation impact (3 citations), the work has already been recognized for its practical deployment potential in programmatic advertising. Agarwal’s research bridges cutting-edge deep learning with real-world fraud detection, offering a scalable, comprehensive defense against one of the internet’s most pervasive threats. His work is essential reading for anyone studying adversarial machine learning or digital advertising security.