Which of the following deployment paradigms can centrally compute predictions for a single record with exceedingly fast results?

Question

Options

• AStreaming
• BBatch
• CEdge/on-device
• DNone of these strategies will accomplish the task.
• EReal-time

• A
  85% (22)
• C
  4% (1)
• D
  4% (1)
• E
  8% (2)

Explanation

Streaming (A) is correct because it processes individual records as they arrive through a data pipeline, computing predictions centrally (on a server) with very low latency - making it ideal for single-record, near-instantaneous inference.

Why the distractors are wrong:

• Batch (B): Processes large groups of records together on a schedule, introducing latency that makes it unsuitable for single-record, immediate results.
• Edge/on-device (C): Computes predictions locally on the end device, not centrally - it violates the "centrally compute" requirement even though it can be fast.
• Real-time (E): While real-time inference is fast, in ML deployment terminology it typically describes a serving pattern (synchronous request/response), whereas streaming specifically describes the data ingestion pattern of processing single records as they flow through a pipeline - making streaming the more precise answer here.
• None of these (D): Eliminated by the existence of streaming.

Memory tip: Think of the word "stream" as water flowing past a single point - one drop (record) at a time, moving fast, processed at a central location. If the question mentions single record + fast + central, stream = the answer.

No community discussion yet for this question.