Google Vertex AI

What it is

Google Cloud's unified ML platform — training, prediction, AutoML Tables, BigQuery ML, Vector Search, plus first-class Gemini API access.

How Vaaani uses it

• Gemini 1.5 Pro for million-token context window
• BigQuery ML for SQL-native model training on warehouse data
• Vertex Vector Search for production semantic retrieval
• Pipelines for MLOps with Kubeflow under the hood

Why it makes the cut

When the customer's data already lives in BigQuery, training models inside the warehouse means zero data movement and faster iteration.

Sample code

from google.cloud import aiplatform
aiplatform.init(project="vaaani", location="asia-south1")

model = aiplatform.Model("projects/.../models/123")
endpoint = model.deploy(machine_type="n1-standard-4")