I am an educator and researcher in computer science and engineering, with over 20 years of university teaching experience and a research program in graph neural networks, graph analytics for financial crime detection, distributed algorithms, computer networks, and applied combinatorics..
My work is driven by a central theme: designing algorithms and graph-based machine learning methods that solve real-world problems on relational data. Graphs arise naturally in communication networks, financial transaction systems, and cybersecurity — in each case, the structure of the graph encodes the key information needed for practical tasks. I design algorithms — both distributed and neural — that extract and exploit this structure for applications such as network optimization, anomaly detection, and financial crime detection.
Graph neural networks for entity resolution and fraud detection. My current research develops the expressivity theory of GNNs for entity resolution and fraud detection. I established tight separation theorems characterizing the minimal message-passing architectures required for different entity-matching tasks on typed entity-attribute graphs, including a sharp complexity gap between detecting any shared attribute and detecting multiple shared attributes — the cross-attribute identity correlation gap. In complementary work, I am designing task-specific, low-complexity GNN architectures for detecting fraud-relevant subgraph patterns, characteristic of money laundering, in directed transaction multigraphs. This work is carried out at the Centre for Cybersecurity, Trust and Reliability (CyStar), CSE department, IIT Madras. [ arXiv:2603.27154 ]
Distributed algorithms for wireless networks. I conducted a research program on performance guarantees of distributed algorithms in wireless networks using graph and hypergraph interference models. I proved tight bounds on the worst-case performance of distributed scheduling algorithms under various interference models, including the primary interference model, the 2-hop interference model, and general hypergraph models. This sole-authored work was published in IEEE/ACM Transactions on Networking, IEEE Transactions on Information Theory, IEEE Transactions on Mobile Computing, Theoretical Computer Science, and Wireless Networks, and addressed open problems in the literature.
Automorphism groups of graphs and interconnection networks. While one can often obtain some automorphisms of a graph, it is generally difficult to prove that one has obtained the full automorphism group. I computed the full automorphism groups of several families of Cayley graphs generated by transposition sets, resolving open problems related to the symmetry structure of interconnection networks and their fault tolerance. This sole-authored work was published in Discrete Mathematics, Journal of Algebraic Combinatorics, and Discrete Applied Mathematics, among other venues.
Across my career, I have contributed to academic program development, faculty recruitment, accreditation, research mentoring, and industry engagement. I have designed curricula for undergraduate and postgraduate programs in CS and AI/ML, delivered corporate training in AI and algorithms, served on institute research boards and faculty interview panels, and prepared accreditation documentation. At one institution, I led a curriculum redesign that grew enrollment sixfold over two years, while similar programs at other institutions were declining or shutting down. Currently at IIT Madras CyStar, I co-authored a funded Dun & Bradstreet CSR grant and provide research guidance to MTech project students and Project Associates working on GNN-based fraud detection and blockchain forensics.
My expertise spans algorithms, graph theory, combinatorics, discrete mathematics, and their applications in computer networks, distributed systems, cybersecurity, and AI for financial crime detection. I combine rigorous mathematical foundations with practical implementation experience. Educated in Austria and the United States, I bring an international perspective on educational quality and curricular rigor to my teaching and curriculum design. Over the years, I have taught more than 3,000 students across undergraduate, postgraduate, and corporate training programs.