Implement knowledge mining and information extraction solutions →hardMultiple Select

AI-102 Practice Question: Implement knowledge mining and information extraction solutions

Which THREE considerations are important when designing a custom skill for Azure AI Search that calls an external API for specialized data extraction?

⚠ Common exam trap

AI-102 often tests the specific numeric limits (16 MB, 230 seconds) and the misconception that custom skills are restricted to a single input/output or a specific programming language.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

The API endpoint must be reachable from the search service

Option A is correct because a custom skill in Azure AI Search is essentially a Web API skill that the indexer invokes over HTTPS, so the external API endpoint must be network-reachable from the search service (including any required private endpoint or firewall rules) or enrichment will fail. Option D is correct because the Web API skill has a documented maximum request payload size of 16 MB; larger payloads will be rejected, so the skill design must keep the serialized input within that limit. Option E is correct because the Web API skill enforces a maximum execution time of 230 seconds per call, after which the request times out, so long-running extraction logic must be optimized or split. Option B is wrong because a custom skill can accept multiple inputs and produce multiple outputs via the inputs and outputs mappings. Option C is wrong because the skill can be implemented in any language or framework (for example C#, Node.js, or Java) as long as it exposes a compatible HTTP endpoint.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✓
The API endpoint must be reachable from the search service
Why this is correct
The search service invokes the custom skill over HTTP during enrichment, so the external API endpoint must be reachable from the service's network path. This satisfies the stem's design consideration, since an unreachable endpoint causes skill execution failures and incomplete indexing.
✗
The skill can only accept one input and produce one output
Why it's wrong here
Custom skills accept multiple inputs and emit multiple outputs, so this restriction is false. Single-input, single-output designs suit trivial transformations, but external extraction skills typically map several enriched fields and return structured results, which the scenario's API-based extraction requires.
✗
The skill must be written in Python
Why it's wrong here
Custom skills run as Azure Functions, which support C#, Java, JavaScript, Python and PowerShell, so Python is not mandatory. Python suits teams already using it for ML processing, but the scenario's external API extraction can be implemented in any supported runtime.
✓
The skill must handle payloads up to 16 MB
Why this is correct
Custom skill inputs and outputs are limited to 16 MB per enrichment payload, so the skill must handle payloads up to that size. This meets the stem's consideration, since exceeding the limit truncates or fails document enrichment during indexing.
✓
The skill must complete within 230 seconds
Why this is correct
Azure AI Search enforces a 230-second execution limit per skill invocation, so the custom skill must complete within that window. This satisfies the stem's consideration, as longer-running external API calls time out and leave documents partially enriched.

Go deeper

AI-102 Practice Question: Implement knowledge mining and information extraction solutions

Why each option matters

Option-by-option breakdown

Related to this question