MLA-C01 Deployment and Orchestration of ML Workflows MEDIUM
PRODUCTION SCENARIO
A risk team scores 40 million account records once each night and writes the results to Amazon S3. No application calls the model between those runs, and the team wants the lowest cost.

Which deployment approach meets the requirement?

Answering here is anonymous. Nothing is saved unless you sign in.

Show answer and explanation

Answer: A batch transform job that partitions the input across instances and then stops

Batch transform partitions the S3 input across instances, writes the output, and terminates, so nothing is billed between nightly runs. A real-time endpoint invoked per record pays for idle hours and adds 40 million round trips. An asynchronous endpoint queues requests one at a time rather than partitioning a large file, and a serverless endpoint caps concurrency far below this volume.
Free

Keep practicing MLA-C01

18 original MLA-C01 practice questions, each with an explanation and a source link. No account needed.

Start free practice set → Timed, explained, free