QuestionQ45
VerifiedSecuring and governing agentic workflowsYou are designing research AI infrastructure for a university computing facility that supports two distinct academic workloads:
- A high-volume, 24/7 data extraction pipeline handling sensitive student records that are subject to strict data residency rules.
- An intermittent, complex research workload requiring advanced multimodal reasoning over large context windows, with unpredictable usage spikes.
You need to minimize total operational expenditure and meet data governance requirements while satisfying the latency and capability needs of both workloads.
What should you do?
Community Discussion