1.2 Compare Model Hosting Options (cloud-Hosted vs Locally Hosted) and Their Trade-Offs (cost, Latency, Privacy, Scalability) AI-TECHNICAL-PRACTITIONER Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   A video streaming service uses an AI model for content moderation. The workload is highly variable, with spikes during live events. They have a dedicated IT team but want to minimize operational overhead. They also need to ensure low latency for real-time moderation. Which hosting strategy is most appropriate?

   • ACloud-host with auto-scaling and a content delivery network (CDN) for edge inference.
   • BLocally host on a small cluster and queue requests during spikes.
   • CCloud-host with a fixed number of instances sized for peak demand.
   • DLocally host on a large cluster sized for peak demand.
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   Which statement accurately describes a typical cost trade-off between cloud-hosted and locally hosted model deployment?

   • ACloud-hosted models typically have higher upfront hardware costs but lower ongoing operational costs compared to locally hosted models.
   • BCloud-hosted models are always more expensive than locally hosted models for any workload.
   • CLocally hosted models typically have higher upfront hardware costs but lower ongoing operational costs compared to cloud-hosted models.
   • DLocally hosted models have no operational costs because the hardware is already purchased.
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A law firm handles highly confidential client documents and wants to use an AI model for document summarization. They have a strict policy that client data must remain within their own data center. Which hosting option is the only one that meets this policy?

   • AUse a hybrid approach with cloud for model training and local for inference.
   • BLocally host the model on servers in the firm's data center.
   • CCloud-host the model with a data residency guarantee in the same country.
   • DCloud-host the model in a private cloud environment dedicated to the firm.
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   Which statement best describes the typical latency trade-off between cloud-hosted and locally hosted model inference?

   • ALocally hosted models generally have lower network latency because data does not need to travel to a remote data center.
   • BCloud-hosted models always have lower latency because cloud providers use faster hardware.
   • CLocally hosted models have higher latency because they lack the cloud provider's optimized network.
   • DLatency is identical for both options because inference time depends only on the model architecture.
   Reveal Solution Discussion Select an answer first
6. 5application · medium

   A financial trading firm needs an AI model for real-time fraud detection. The model must respond in under 10 milliseconds. The firm has a stable, predictable workload and a large IT budget. Which hosting option is most appropriate?

   • ACloud-host the model with a dedicated private connection.
   • BLocally host the model on dedicated high-performance servers.
   • CCloud-host the model in a region with low latency to the firm's network.
   • DUse a hybrid approach with local inference and cloud fallback.
   Reveal Solution Discussion Select an answer first