You are a lead ML architect at a small company that is migrating from on-premises to Google Cloud. Your company has limited resources and expertise in cloud infrastructure. You want to serve your…

Question

Options

• AConfigure Compute Engine VMs to host your models.
• BCreate a Cloud Run function to deploy your models as serverless functions.
• CCreate a managed cluster on Google Kubernetes Engine (GKE), and deploy your models as
• DDeploy your models on Vertex AI endpoints.

• A
  5% (3)
• B
  2% (1)
• C
  11% (6)
• D
  82% (46)

Explanation

The scenario emphasizes limited cloud expertise, wanting to serve models quickly, with scalability, reliability, and no additional resource management. Vertex AI endpoints are a fully managed serving infrastructure: you register a model, create an endpoint, and deploy-Google handles autoscaling, load balancing, health checks, and infrastructure. Option A (Compute Engine VMs) requires provisioning, configuring, and maintaining virtual machines. Option C (GKE) requires managing Kubernetes clusters, which is complex and resource-intensive. Option B (Cloud Run functions) is serverless but has memory/size limitations and is not optimized for ML model serving the way Vertex AI endpoints are. Vertex AI endpoints are the purpose-built, zero-infrastructure-management ML serving solution on Google Cloud.

No community discussion yet for this question.