Intro to GCP for Machine Learning & AI

Workshops
Code-along
Carpentries
Compute
Cloud
Google
GCP
GPU
LLM
RAG
Retrieval
Author

Chris Endemann

Published

March 5, 2026

This Intro to GCP workshop teaches core workflows for building, training, and tuning ML/AI models in Google Cloud’s Vertex AI platform. Participants learn to set up data, configure Vertex AI Workbench notebooks, launch training and tuning jobs, and optimize resource costs effectively within GCP. The workshop also includes a section on building retrieval-augmented generation (RAG) pipelines using Gemini models.

TipUW-Madison Cloud Users

A personal GCP account is fine for this workshop. However, for long-term research use, we recommend switching to a UW-provisioned GCP account. You’ll get institutional pricing, lower overhead on grants (26% instead of 55.5% — saving ~$2,950 per $10k in cloud costs), data protection agreements (including BAA for HIPAA), and dedicated support from the Public Cloud Team. NIH-funded researchers can get additional discounts through the STRIDES Initiative. You can also apply for $5,000 in Google Cloud Research Credits.

Request a UW GCP account | Why use a UW account? | Full details: UW Cloud Services

Cost estimate

Running through this workshop should cost approximately $3–$8 on GCP, assuming short GPU runs and limited hyperparameter tuning trials. Using n2-standard-4 or e2-standard-4 instances with a single T4 GPU generally stays within this range. New accounts may be eligible for $300 in free GCP credits, which typically cover the full cost of this workshop. It is recommended to track usage in the GCP Billing Console and delete unused resources once completed.

Prerequisites

Estimated time to complete

4–6 hours: Based on running through training, tuning, and the Gemini RAG pipeline example.

Comments