Dr. Sajal Saha
University of Northern British Columbia (UNBC), Canada.
Department of Computer Science
Title: Building AI-Driven Cyber Defense Agents: From Theory to Practice
Large language models (LLMs) are opening up new possibilities in cybersecurity, from
detecting threats to generating real-time responses. This tutorial explores how LLMs—
when
combined with the LangChain framework—can be used to build intelligent and modular cyber
defense agents. We start with the basics of using LLMs in cybersecurity,
including prompt design,
chaining logic, memory, and interactive reasoning. Then, we show how LangChain can bring these
pieces together to create systems that detect attacks, explain decisions, and suggest mitigation
steps. We focus on building pipelines that use retrieval-augmented generation (RAG) to pull in
threat intelligence and logs, helping
LLMs respond with context-aware answers. We also highlight
how explainable AI can surface key indicators of attacks and guide response generation. Finally,
we discuss
how to integrate external tools and APIs for automated mitigation.
Biography
Sajal Saha is an Assistant Professor in the Department of Computer Science at the University
of Northern British Columbia (UNBC), Canada, where he leads the INFORM Lab.
His research
focuses on Internet traffic analysis, cyber-attack detection and mitigation, and privacy-preserving
machine learning, with applications in building intelligent and secure network systems. He
has
authored over 30 peer-reviewed publications, including papers in IEEE Transactions and top-tier
conferences such as ICC, INFOCOM, CCNC, and NOMS. Dr. Saha earned his Ph.D. from Western
University and
M.Sc. from Brock University, where he also held research and teaching roles. He
has contributed to academic service in various capacities, including serving as a session chair at
IEEE ICC and as a TPC member
for several international conferences such as IEEE GLOBECOM.
He is actively engaged in research collaborations with NSERC, Mitacs, and industry partners,
focusing on the development of robust, end-to-end frameworks for cyber-attack detection and
autonomous mitigation.
Dr. Lilatul Ferdouse
Wilfried Laurier University, Waterloo, Ontario, Canada
Department of Physics and Computer Science ,
Title: Intelligent Meta-Surfaces Assisted Unmanned Aerial Vehicle Networks: Revolutionizing 6G and Beyond
Unlocking the connectivity and efficiency of 6G networks involves exploring cutting-edge technologies like meta-surface.Meta-surfaces, also known as intelligent reflecting
surfaces (IRS), are metamaterials engineered with properties not found in naturally occurring materials. Reflecting on their potential, these surfaces could revolutionize wireless communication by manipulating electromagnetic waves in novel ways. IRS-assisted Unmanned Aerial Vehicle (UAV) networks are emerging as a promising solution to enhance
wireless communication coverage, reliability, and spectral efficiency in dynamic environments. By integrating IRS technology with UAV platforms, these networks can intelligently reconfigure the wireless propagation environment to support flexible deployment, energy-efficient transmissions, and improved user connectivity. This talk provides a concise overview of the key benefits,
challenges, and applications of IRS-assisted UAV networks, highlighting their potential in next-generation wireless systems such as disaster recovery, remote sensing, and mobile edge computing.
Biography
Dr. Lilatul Ferdouse received her PhD and MASc in Electrical and Computer Engineering from Ryerson University in 2019 and 2015, respectively. Prior to joining Wilfried Laurier University, She was a
postdoctoral fellow at DABNEL lab, Department of Computer Science at Ryerson University. In 2020, she also worked as a research assistant at the Department of Computer Science at Thompson River University, Kamloops, BC, Canada. Her research interests are in resource allocation of next generation wireless communication systems, relay-assisted cooperative networks, congestion and overload control techniques in machine-to-machine and industrial IoT networks.