Bryan Perozzi

(Research Scientist, Google Research)

Bio: Bryan Perozzi is a Research Scientist in Google Research, where he routinely analyzes some of the world’s largest (and perhaps most interesting) graphs. Bryan’s research focuses on developing techniques for learning expressive representations of relational data with neural networks. These scalable algorithms are useful for prediction tasks (classification/regression), pattern discovery, and anomaly detection in large networked data sets. Bryan is an author of 40+ peer-reviewed papers at leading conferences in machine learning and data mining (such as NeurIPS, ICML, ICLR, KDD, and WWW). His doctoral work on learning network representations (DeepWalk) was awarded the prestigious SIGKDD Dissertation Award. Bryan received his Ph.D. in Computer Science from Stony Brook University in 2016, and his M.S. from the Johns Hopkins University in 2011.

Title of Talk: A Few Foundational Takes on Connected Data

Abstract: Graphs are powerful tools for representing complex real-world relationships, essential for tasks like analyzing social networks or identifying financial trends. While large language models (LLMs) have revolutionized natural text reasoning, their application to graphs remains an understudied frontier. In this talk, I’ll describe some highlights from our journey making LLMs more relational, and then focus on our new work making graph models more foundational. I’ll introduce the Universal Classifier (UC), which supports arbitrary feature and class cardinalities spanning all learning objectives through the in-context learning paradigm. The UC reformulates all node-, edge-, and graph-level prediction tasks as maximizing similarity in the latent space: by lifting heterogeneous features and labels into 3D latent tensors, the model learns transferable features independent of specific input schemas. This architecture allows a single pre-trained model to generalize to node classification, node regression, and link prediction across unseen graphs with varying feature semantics. Experiments show strong zero-shot transfer performance across node- and link-level tasks.

Vatche Isahagian

(Senior Research Scientist, IBM Research)

Bio: Vatche Isahagian is a Senior Research Scientist and manager at IBM Research in Cambridge, Massachusetts. He earned his PhD in Computer Science from Boston University. His research spans a broad set of disciplines across distributed systems, Artificial Intelligence, and business processes. His current focus is on utilizing AI techniques such as natural language processing and multi-agent systems to enable AI-enhanced business automations. Vatche’s work has resulted in multiple patent filings, peer-reviewed publications in conferences and journals, as well as two best paper awards. Additionally, he has organized several workshops and served as a member of program committees, as well as co-chair and publicity chair for various conferences. He is a senior member of both the IEEE and the ACM.

Jayaram K Radhakrishnan

(Senior Staff Research Scientist, IBM Research)

Bio: Jayaram is a researcher and senior technical leader at IBM Research AI. He leads projects and efforts on applied Generative AI and federated learning, especially on copilots and agents for business automation. He also works at the intersection of confidential computing and AI, in the context of confidential model inference and federated learning.
Jayaram also has broad expertise in distributed systems, middleware and cloud computing, including systems for training AI models, event-based systems and stream processing. He has won several IBM awards, and chairs an IBM-wide evaluation committee on patent filings in the cloud computing area. He contributes to the research community through service on various conference technical program committees, and has served as conference chair at MIDDLEWARE, ICDCS, MASCOTS and other conferences.