Purdue University MSAI Virtual Speaker Series Proposal Proposed Topic: Managing AI Cost at Scale: From Model Training to Production Inference

Created using ChatSlide

This presentation explores AI cost management by balancing expenses and performance. It covers key cost drivers such as GPU compute, strategies for optimizing training and inference costs, and financial best practices including FinOps and cost KPIs. Techniques like mixed precision, spot instances, model compression, and caching are discussed alongside real-world case studies to achieve cost-effective AI scalability.