A company has an ML model that needs to run one time each night to predict stock values. The model input is 3 MB of data that is collected during the current day. The model produces the predictions…

Question

Options

• AUse a multi-model serverless endpoint. Enable caching.
• BUse an asynchronous inference endpoint. Set the InitialInstanceCount parameter to 0.
• CUse a real-time endpoint. Configure an auto scaling policy to scale the model to 0 when the
• DUse a serverless inference endpoint. Set the MaxConcurrency parameter to 1.

• A
  12% (5)
• B
  5% (2)
• C
  2% (1)
• D
  81% (35)

Explanation

Option D is correct because SageMaker Serverless Inference is purpose-built for infrequent, sporadic workloads - you pay only when invoked, making it ideal for a once-nightly job. The 3 MB payload falls within serverless limits (6 MB max), and the sub-1-minute runtime fits well within execution constraints. Setting MaxConcurrency=1 reflects that only one prediction job runs at a time.

Option A is wrong because a multi-model endpoint is designed to host multiple models on shared infrastructure - overkill for a single model. Caching also provides no benefit when the workload runs only once per night with different daily data.

Option B is wrong because asynchronous inference is designed for large payloads (up to 1 GB) and long-running jobs (up to 1 hour). This job is neither - 3 MB and under a minute doesn't justify async. Also, InitialInstanceCount=0 is not a valid parameter for achieving scale-to-zero on async endpoints without additional configuration.

Option C is wrong because SageMaker real-time endpoints cannot scale to zero - they always maintain at least one running instance, meaning you'd pay for idle compute all day when the model only runs at night.

Memory tip: Match the deployment type to the usage pattern - "once a night" = infrequent = Serverless (pay-per-invocation). Real-time = always-on cost; Async = big/slow jobs; Multi-model = many models, one endpoint.

No community discussion yet for this question.