Session: Democratizing Large models for Edge Inferencing
As models grow larger and more powerful, running them on devices like smartphones, cameras, and IoT sensors becomes increasingly challenging. Democratizing Large Models for Edge Inferencing seeks to make these models more accessible by enabling them to run directly on smaller, resource-constrained devices. This reduces the need for costly GPU servers, offering faster, more secure, and cost-effective solutions. The topic explores model compression techniques that make large models feasible on the edge, while also highlighting their potential impact across industries. By bringing complex processing closer to users, this approach ensures better data privacy and real-time performance.
Bio
Dhivya is an accomplished AI/ML practitioner with over 15 years of experience in the finance industry. Throughout her career, she has held diverse roles across organizations, always driven by a deep curiosity and a passion for harnessing the full potential of data to drive business value. In her current role as Lead AI Solutions Architect, she is leading critical Generative AI initiatives for the bank, with a focus on solutions that have the potential to achieve significant cost savings.
Her primary area of research is in democratizing AI for Edge Inferencing, striving to make advanced AI models more accessible and efficient across resource-constrained environments.
In addition to her professional work, Dhivya has been an active mentor for prominent events such as TreeHacks, Stanford’s premier hackathon, and the Analytics for Good Hackathon.
She has been invited to speak at major industry conferences, including the Data Science Salon in Austin and The Data Science Conference in Chicago.
Dhivya is a member of the AI Frontier Network, an exclusive, invite-only organization, and serves on the advisory board for the University of Minnesota’s graduate analytics program. She is also a Women Executive Leader member of the Forbes Technology Council.
She holds a Master’s in Business Analytics and a Bachelor’s in Information Technology, along with a Graduate Certificate in Business Communications from Harvard University. Dhivya is currently based in Minnesota.