Back to feed
arXiv cs.AI·

Evaluating Temporal Semantic Caching and Workflow Optimization in Agentic Plan-Execute Pipelines

Signal
75
Hype
15
In three linesOptimization study for industrial agent pipelines (AssetOpsBench). Proposes temporal semantic cache and MCP optimizations (tool discovery, parallel execution): 1.67x speedup, 40% latency reduction, 30.6x on cache hits. Identifies failure modes of pure semantic caching for parameter-rich queries.
Read source
Your take?
AI AgentsMCPBenchmarksEvals

Summary generated by Claude — human-verified