Preprint
Aug 2026
ScaleSense: Cost-Intelligent Scaling Framework via Learned Resource Estimation in Alibaba AnalyticDB
This work proposes ScaleSense, a proactive, query-level resource scaling framework that addresses the critical performance-cost trade-off while maintaining low-overhead inference latency, and confirms its practical performance in production deployments.
Yifan Wu, Yuhan Li, Zhenhua Wang et al.
· 0 citations