Skip to content

Quickstart without infrastructure

You don't need any infrastructure to start with Polars Nanook. Install the client and sign up, and the queries you run on your own machine show up in the portal with a full profile. Invite your team to the same workspace to share them. When a query outgrows your machine, or you want Polars Nanook to start machines for you, connect a deployment in your infrastructure.

How do I install the client?

Install the open-source Polars library and the polars-nanook package in your environment:

pip install polars polars-nanook

The package includes the nanook command line tool.

How do I sign up?

Sign up at nanook.pola.rs, or run nanook authenticate, which opens the same page in your browser. Polars Nanook creates a default organization and workspace for you. Set up an organization lists their names and how to rename them.

The portal opens on a welcome page with three steps: profile a query, connect your infrastructure and invite your team. Complete them in any order. The Get started! card in the sidebar tracks your progress.

Welcome page showing the three-step checklist: profile a query from your laptop, connect your own infrastructure, invite your team

How do I profile a query on my own machine?

Add pl.Config.enable_monitoring() above a lazy query and run it as usual. This example runs on its own:

import polars as pl
import polars_nanook as pn

pn.authenticate()

pl.Config.enable_monitoring()

print(
    pl.LazyFrame(
        {
            "quantity": [1, 2, 3, 4, 3],
            "price": [34.2, 43.2, 123.1, 23.3, 42],
            "order_id": [1, 1, 2, 2, 2],
        }
    )
    .group_by("order_id")
    .agg((pl.col("quantity") * pl.col("price")).sum())
    .collect()
)

The query still runs locally. Polars sends the query plan and per-operator runtime stats to your workspace, where the profiler shows which step is slow. Open the query from the Queries tab of your workspace to see the profile.

pl.Config.enable_monitoring(workspace=...) sends profiles to a workspace other than your default. Monitoring runs queries on the streaming engine, which collects the per-operator metrics.

How do I invite my team?

An organization admin invites colleagues to the workspace by email from the Invite your team card on the welcome page or from the Team page in the sidebar. While the workspace still has its default name, the dialog first asks you to rename it.

Rename your workspace step of the invite dialog

Invite a member step of the invite dialog with the email field

Members can inspect every query profile in the workspace, including profiles of local queries other members sent with enable_monitoring(). See the team page for workspace roles and organization members for access at the organization level.

How do I run queries in my infrastructure?

Connect a deployment to the workspace, and remote queries run on compute in your infrastructure. Choose a deployment compares the deployments.

  • AWS


    Polars Nanook starts EC2 instances in your AWS account on demand, through a CloudFormation stack you deploy once.

    Get started on AWS

  • Kubernetes Helm chart


    A cluster you run on EKS, GKE, AKS or any CNCF-conformant distribution, installed with the Helm chart.

    Get started with the Helm chart

  • Bare-metal


    A cluster you run on your own servers, from the polars-nanook-compute package and one TOML file.

    Get started on bare-metal

One-time workspace setup

You connect a deployment once per workspace. Members you invite to a connected workspace can run remote queries right away, without any setup of their own.

Next steps