Sign In
Register

Partnership opportunities

Tickets

Call to action
Your text goes here. Insert your content, thoughts, or information in this space.
Button

Back to speakers

Junwei
Huang
Director, Data Science
Mastercard
Junwei is a recognized voice at the intersection of Generative AI, financial technology,and enterprise-scale data systems. As Director of Data Science at Mastercard, he leads a global team advancing the frontiers of transaction intelligence, applying large language models and generative systems to enrich merchant data and mitigate disputes across billions of transactions. His team’s innovations directly support $86M in revenue and shape how AI is deployed responsibly in financial services. With over 15 years of experience spanning academia, industry, and government research, Junwei brings deep technical expertise in LLMs, deep learning, reinforcement learning, and MLOps. His prior roles include leading AI development at RBC, Network Analytics at Scotiabank, as well as algorithmic research for Natural Resources Canada. He has applied AI in diverse domains—from entity resolution in e-commerce to multi- modal model based content moderation and LLMs-based subscription detection. An adjunct professor at Northeastern University, Junwei teaches advanced GenAI and data science courses and mentors the next generation of AI talent. He also actively contributes to the AI community through workshops, advisory roles, and mentoring programs. Junwei holds a Ph.D. from the University of Toronto.
Button
20 November 2024 12:00 - 12:30
Panel | From pilot to production: Shipping enterprise ready AI systems
In this session, we’ll walk through the architectural, operational, and governance decisions that define enterprise-ready GenAI systems, including: Key takeaways: → Designing for real-time performance, throughput, and latency across use cases. → Choosing between proprietary models, open source, and fine-tuned LLMs and managing that complexity over time. → Building observability and feedback loops into your stack to track model drift and measure business impact. → Putting the right controls in place for data privacy, access, and regulatory compliance.