RAM GANESH
Senior Data Scientist at Mastercard
Ram is a Senior Data Scientist, in the AI Research Team of Mastercard specializing in large-scale graph representation learning, Generative AI, fraud detection and anti-money laundering. At Mastercard, he develops AI-driven solutions that serve both internal and external stakeholders, generating multi-million-dollar revenue impacts. A graduate of IIT Madras, Ram cleared CFA L3 examination and demonstrates profound deep technical and financial expertise. His research has resulted in numerous patents and research paper publications, especially in AI for financial systems and large-scale graph learning. He currently works on enhancing reasoning capabilities of foundation models and scalable graph learning. Ram has played a crucial role in organizing and hosting high-profile conferences and technical sessions within his areas of expertise, such as the INAUGURAL Learning on Graphs Conference (LoG) New Delhi Meetup in South Asia. He also leads various strategic collaboration initiatives with leading academic institutions, aiming to bridge the gap between industry and academia. An advocate for diversity in technology, Ram also actively supports Women in Tech initiatives, further emphasizing his commitment to inclusive work environments.
May 13th-14th, 2025
Smaller, Smarter, Sustainable: Optimizing LLM's by Leveraging White Box Knowledge Distillation Algorithms
LLM model sizes have been increasing exponentially with GPT 4 having more than 1T parameters at least. However, the associated computational and memory requirements of such models might hinder widespread adoption. To address this, knowledge distillation (KD) has become essential for transferring the capabilities of massive models into smaller, more efficient counterparts for various downstream tasks Beyond reducing size, KD enhances environmental sustainability and fosters an equitable AI landscape by making advanced capabilities accessible to smaller entities.
Brought to you by
