Рет қаралды 53
This webinar will provide engineering managers with crucial strategies for scaling AI/ML infrastructure. Hope Wang will share her expertise from working with industry giants like Uber and Expedia Group, focusing on overcoming challenges related to cost, complexity, and scalability. Attendees will gain actionable insights on enhancing performance, managing cloud and network expenses, and optimizing GPU utilization to drive more effective and efficient engineering practices.
Lessons Learned:
Overcoming Data Locality Issues: Learn from Uber's experience with Alluxio in managing multi-region/cloud data locality challenges to reduce operational overhead and latency, enhancing system responsiveness.
Enhancing Model Training Efficiency: See how Uber accelerates model training by integrating cutting-edge technologies to reduce data load times and boost GPU efficiency, streamlining workflow for engineering teams.
Cost-Effective Data Management: Gain insights from Expedia Group's approach to managing extensive datasets across various storage systems and clouds, minimizing costs through strategic data access and replication tactics.
Streamlining Data Access and Integration: Understand the integrative data strategies these tech leaders employ for seamless access across diverse data sources, effectively minimizing I/O bottlenecks and improving system performance.
Adopting Cloud-Native Solutions: Discover the benefits of cloud-native solutions for distributed data management, as emphasized by top industry players, to support robust analytics and AI applications at scale.
Navigating Data Engineering Complexities: Learn practical solutions for managing complex data engineering challenges, facilitating faster AI development and deployment, and enhancing team productivity.