Vikas C. Gangadhara

Behavioral Intelligence for Detecting Bots and Automated Abuse in Mobile Applications

Abstract:

Mobile bot detection is becoming increasingly challenging as attackers evolve beyond conventional scripts to sophisticated techniques involving emulators, instrumented devices, replay tools, device farms, synthetic identities, and automated application workflows. This talk presents behavioral intelligence as a mobile-native approach for determining whether an end-to- end interaction is consistent with legitimate human behavior, rather than relying solely on isolated identifiers, device attributes, or individual requests. The proposed framework integrates interaction, device-context, runtime, sensor, network, session, workflow, and backend telemetry, transforming raw signals into temporal sequences, behavioral embeddings, user-flow graphs, and contextual features. Detection is formulated as a sequential intelligence problem that combines anomaly detection, probabilistic risk scoring, sequence modeling, graph learning, and unsupervised and semi-supervised techniques to identify both known and previously unseen forms of automation. Attention is given to touch dynamics, navigation patterns, device motion, runtime integrity, API sequencing, and inconsistencies between client-side actions and backend events. The architecture also separates detection from response, enabling adaptive interventions such as enhanced monitoring, throttling, step-up verification, or blocking based on confidence levels and workflow sensitivity. The talk further addresses false-positive reduction, adversarial robustness, privacy-aware instrumentation, and continuous learning, while exploring emerging directions such as multimodal behavioral models, graph-based reasoning, privacy-preserving analytics, and autonomous security agents for adaptive defense against rapidly evolving mobile application abuse.

Profile:

Vikas C. Gangadhara is a technology and engineering leader with more than 18 years of experience in mobile application architecture, cybersecurity, cloud technologies, distributed systems, and enterprise software development. His expertise spans iOS, Android, web technologies, AWS cloud services, networking, and secure software architecture. Vikas currently serves in a principal engineering leadership role at Akamai Technologies, where he contributes to the development and advancement of Akamai Bot Manager Premier (BMP). His work focuses on protecting mobile and web applications from sophisticated automated threats through secure telemetry collection, mobile SDK development, real-time detection systems, and advanced bot-detection techniques. He also leads cross-functional engineering initiatives spanning architecture, threat modeling, technical strategy, engineering execution, and product delivery. Vikas is an inventor in the cybersecurity domain and is named on a U.S. patent- approved invention related to dynamic data-signal collection designed to prevent telemetry spoofing in bot-detection systems. Previously, he held senior engineering and architecture roles at TaxAct, Genpact, Endeavour Software Technologies, and SPAN InfoTech, contributing to solutions across mobile banking, tax technology, home automation, networking, and enterprise mobile applications. His technical expertise includes Swift, Objective-C, Kotlin, Java, JavaScript, Python, C++, AWS, microservices, APIs, and networking technologies. Vikas holds a Bachelor of Engineering in Electronics and Communication Engineering and is an AWS Certified Cloud Practitioner. His career reflects a strong combination of hands-on engineering expertise, technical leadership, cybersecurity innovation, and the delivery of secure and scalable technology solutions.