Agentic AI on Edge
Building Hybrid AI Systems: Edge Inferencing Meets Cloud Reasoning
On October 17, talks finish at 12:15. The lunch break runs from 12:15 to 14:00; the dining location will be announced.
Talk overview
Using warehouse safety monitoring as a case study, this talk presents a hybrid architecture that runs latency-sensitive vision-language inference on edge GPUs while delegating summarization and conversational retrieval to cloud agents. It covers workload placement, model selection, optimization, shared memory, and cross-environment observability.
