PMLE Serving and Scaling Models Practice Question

Answer analysis

Option-by-option breakdown

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

• ✗

  Vertex AI AutoML

  Why it's wrong here

  AutoML trains custom classification, regression, image and text models from labelled data; it neither constructs nor manages ANN indexes for similarity search. It is tempting because AutoML also produces deployable models, yet nearest-neighbour retrieval at scale requires Vertex AI Vector Search, which handles index building and low-latency querying.

• ✗

  Vertex AI Workbench

  Why it's wrong here

  Workbench provides managed JupyterLab notebook environments for data exploration and model development; it hosts no ANN index service. It is tempting because notebooks can call embedding models and query indexes interactively, but building and managing approximate nearest neighbour indexes at scale is the role of Vertex AI Vector Search.

• ✗

  Vertex AI Prediction

  Why it's wrong here

  Vertex AI Prediction serves deployed models for online or batch inference; it does not build or manage ANN indexes. It is tempting because similarity search queries ultimately run against an endpoint, but the index construction, updates and serving of nearest-neighbour matches belong to Vertex AI Vector Search.

• ✓

  Vertex AI Matching Engine (Vector Search)

  Why this is correct

  Matching Engine (Vector Search) is the Vertex AI service purpose-built for creating and serving ANN indexes, delivering scalable similarity search across billions of embeddings. It directly satisfies the stem's requirement for approximate nearest neighbour indexing at scale, unlike general prediction or training services.

Go deeper

Related to this question