mediumMultiple Choice
Generative AI Leader Practice Question: A data science team wants to build a RAG pipeline…
A data science team wants to build a RAG pipeline to ground a chatbot in proprietary knowledge. They need to choose a vector database and embedding model. Which combination is NATIVELY integrated with Vertex AI and requires the least custom infrastructure?
⚠ Common exam trap
Google Cloud exams often test the distinction between 'natively integrated' and 'compatible' — candidates may assume any popular vector database like Pinecone works seamlessly with Vertex AI, but only Vertex AI Vector Search offers native, infrastructure-free integration.
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
✓
Vertex AI Embeddings API + Vertex AI Vector Search
Vertex AI Embeddings API and Vertex AI Vector Search are both native, fully managed services within the Vertex AI ecosystem, requiring zero custom infrastructure for deployment. The Embeddings API generates text embeddings directly, and Vector Search provides a scalable, low-latency vector database that integrates seamlessly without additional servers or third-party tools.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
TensorFlow Embedding Projector + BigQuery
Why it's wrong here
Embedding Projector visualises high-dimensional embeddings for analysis; it neither generates embeddings nor performs similarity search, and BigQuery lacks native vector indexing for retrieval. It fits exploratory embedding inspection, not a production RAG pipeline requiring low-latency nearest-neighbour queries.
- ✓
Vertex AI Embeddings API + Vertex AI Vector Search
Why this is correct
Both Vertex AI Embeddings API and Vertex AI Vector Search are first-party Vertex AI services sharing native integration, authentication and data flow, eliminating the custom glue code and infrastructure that third-party vector databases would require.
- ✗
Custom embeddings using a BERT model + Elasticsearch
Why it's wrong here
Hand-rolled BERT embeddings plus Elasticsearch means training, serving and index plumbing the team must build and maintain themselves, adding custom infrastructure. That approach suits teams needing bespoke domain embeddings with existing Elasticsearch expertise, not a natively integrated managed stack.
- ✗
Vertex AI Embeddings API + Pinecone
Why it's wrong here
Pinecone is a third-party vector store, so wiring it to Vertex AI needs custom connectors, credential handling and index synchronisation rather than native managed integration. It suits teams already standardised on Pinecone across clouds, but here Vertex AI Vector Search removes that infrastructure burden.
Go deeper
Related to this question
About these practice questions
Courseiva writes every Generative AI Leader question from scratch — 1,008 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This Generative AI Leader practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the Generative AI Leader exam.