GCP Basics · Module 10

AI on GCP – Gemini, RAG, BigQuery ML and Agents

Find your way around Google’s AI platform (Vertex AI, now part of Gemini Enterprise Agent Platform), call Gemini models securely with IAM and quotas in mind, ground answers in your own data with RAG and vector search, run ML and generative AI directly in BigQuery, and deploy agents with the right guardrails.

Practice interactively →

What you will learn

• Explain the parts of Vertex AI / Gemini Enterprise Agent Platform and when to use each
• Call a Gemini model with IAM, Application Default Credentials, quotas and safety settings
• Design a retrieval-augmented generation (RAG) flow with embeddings and vector search
• Use BigQuery ML and AI functions from SQL
• Deploy an agent with least privilege, Model Armor and evaluation

Lessons

1. Vertex AI and Gemini Enterprise Agent Platform: the Map

   Vertex AI, now delivered as Gemini Enterprise Agent Platform, groups models (Model Garden), tools to build and test prompts and agents, runtime for agents, and MLOps for custom models; most teams start by calling a hosted model.

2. Calling Gemini: Auth, SDK, Quotas, Cost and Safety

   Call Gemini with the Google Gen AI SDK or REST using Application Default Credentials and the Vertex AI User role; plan for tokens, quotas and 429s, and use safety settings and Model Armor to filter harmful input and output.

3. RAG on GCP: Embeddings, Vector Search and RAG Engine

   Retrieval-augmented generation finds relevant chunks of your documents with embeddings and vector search, then gives them to the model so answers are grounded and cite sources; Google Cloud offers managed and do-it-yourself options.

4. BigQuery ML and AI Functions in SQL

   BigQuery ML trains and runs models with SQL, and remote models plus AI functions call Gemini and embedding models directly from queries through a BigQuery connection.

5. Agents on Google Cloud and AI Security

   Agents combine a model with tools and memory; build them with the Agent Development Kit, host them on Agent Runtime or Cloud Run, and protect them with least-privilege identities, Model Armor, VPC Service Controls, logging and evaluation.