Skip to main content

Working with Quanton locally

A guided, copy-paste run through Quanton on minikube. By the end you'll have a working local Quanton deployment, a real benchmark comparing OSS Spark to Quanton, and an LLM-backed agent reasoning about a live Spark job.

Time required: about 20–30 minutes, most of it waiting on data generation and benchmark queries.

What you'll do

  1. Setup your local environment — minikube + CLI installs
  2. Exercise 1 — run your first QKO example (Pi)
  3. Exercise 2 — use Agentic AI on a live Spark job
  4. Exercise 3 — compare Quanton vs OSS Spark on TPC-DS

Use the Next button at the bottom of each page to advance.

Prerequisite

This workshop assumes you've completed Setup the Onehouse project — you'll need the onehouse-values.yaml from that step to install the operator.