Mr. Soumyajyoti Bhattacharya

Beyond the Profile: Evolving Personalization Through Multi-Signal Intelligence

Abstract:

Personalization is evolving from tracking user actions to interpreting the intent behind them. Modern AI systems increasingly rely on multiple layers of signals that extend beyond the individual user, incorporating marketplace trends, product semantics, and contextual intent. This session examines how multi-encoder architectures, such as two-tower and transformer-based models, can learn from these diverse inputs to create deeper and more adaptive personalization. Drawing from real-world applications in fintech and commerce, it highlights how integrating non-user signals improves discovery, addresses cold start challenges, and builds fairness. The talk envisions personalization that understands context, not just behavior.
 

Profile:

Soumyajyoti Bhattacharya is a Senior Software Engineer at PayPal, where he designs and deploys large-scale intelligent systems that power personalized experiences across PayPal, Venmo, and Honey. His work spans AI-driven personalization, product ranking models, and embedding pipelines—combining behavioral, contextual, and semantic signals to build adaptive commerce ecosystems serving millions of users. Before rejoining PayPal, he engineered data-intensive APIs and content delivery systems at Warner Bros. Discovery and Expedia, bringing efficiency, scalability, and intelligence to global products. His experience bridges fintech, entertainment, and travel technology, with a consistent focus on transforming data into meaningful and sustainable user value.
Soumyajyoti is also pursuing a PhD in Information Technology, researching the ethics and security of intelligent systems, including fairness in agentic workflows and governance frameworks for responsible AI. Beyond his professional work, he serves on nonprofit boards, supports mental health initiatives, and mentors early-career technologists through IEEE and major hackathons.

His public speaking and research contributions include appearances at TEDx, Major international Research Conferences, and global innovation forums, where he continuously advocates for an inclusive, transparent, and human-centered AI future.