1. Welcome to Groundzero

Getting StartedDownload PDF (12.0 MB)
Welcome to GroundzeroA quick tour: home chat, navigation and search.

Groundzero is a data platform that runs in your browser. You use it to bring your data together, understand it, turn it into charts and dashboards, keep it up to date automatically, and ask an AI assistant for help along the way.

This guide is your first hour with Groundzero. By the end you will have connected a data source, explored it, built a dataset, drawn a chart, published a dashboard, scheduled a job and checked where your data comes from.

What you can do with Groundzero#

  • Connect the systems that hold your data, such as a database, a data warehouse or cloud storage.
  • Find and understand your data in a searchable catalog, with descriptions and tags.
  • Ask questions of your data and save the answers as reusable datasets.
  • Show the answers as charts and dashboards that you can share.
  • Automate data preparation with jobs that run on a schedule, and explore with notebooks.
  • Trace where every table comes from and what depends on it.
  • Ask for help from an AI assistant that knows Groundzero.

A map of the main areas#

Seven places do most of the work. Six sit in the bar at the top of every page, and the AI assistant is the home page.

Area Use it to You will use it in
Connections Save links to your data sources Step 1
Catalog Find data, read and write descriptions, see lineage Steps 2 and 7
Designer Build datasets, filters, visualizations and dashboards Steps 3, 4 and 5
Workspaces Share published dashboards in named folders, and browse your storage buckets (My Files) Step 5
DE/ML Keep jobs, notebooks, models and agents in projects, and schedule jobs Step 6 and chapter 4
Lakehouse Store governed tables in warehouses and query them with SQL Chapter 6
Home (AI assistant) Ask questions in plain language and get help writing code Chapter 5

Most areas also have a narrow menu on the left. Widen it with the arrow at the bottom left of the page.

The building blocks#

Groundzero uses a small set of words. Each one is explained here the first time it appears, and again in the glossary at the end.

BRING DATA IN Connections saved links to sources Lakehouse warehouses of tables SHAPE AND SHOW Datasets saved SQL or code Visualizations 22 chart types Dashboards published to Workspaces BUILD AND RUN Project holds your engineering work ETL jobs Notebooks Models AI agents Runtime (RTE) Standard, DuckDB, Spark UNDERSTAND AND ASK Catalog and lineage find, describe, trace Ontology Studio shared business definitions Groundzero AI assistant chat, code help, approvals
Figure 1. The Groundzero building blocks. Connections feed datasets, jobs and notebooks; runtimes run your code; the catalog and lineage describe what exists.
Word What it is
Site Your organisation's Groundzero. Your administrator chooses its logo, colours and sign-in background.
Connection A saved link to an outside system such as PostgreSQL, Snowflake, Amazon S3, a REST API, an AI provider or a Groundzero Lakehouse. Your code and datasets use the connection's name and never contain a password.
Lakehouse and warehouse A Groundzero Lakehouse stores tables in a warehouse. A warehouse holds databases, tables and columns that you can query with SQL.
Dataset A saved, named result. You define it with a SQL query (or Python code) on one or more connections. Charts and dashboards are built on datasets.
Visualization A saved chart or table built on a dataset. Groundzero offers 22 chart types.
Dashboard A page of visualizations, arranged in tabs, with filters. You can publish a dashboard and import it into a workspace folder so others can read it.
Project A container in DE/ML for your ETL jobs, notebooks, models and AI agents.
ETL job Python code that reads data from connections, changes it and writes the result. It runs when you ask or on a schedule. (ETL means extract, transform, load.)
Notebook An interactive document of code and text cells. You run it in a session and see results as you go.
Runtime (RTE) and compute The runtime is the environment that runs your code (Standard, DuckDB or Spark). Compute is the memory and processor size you give it.
Catalog A searchable list of datasets, ETL jobs and warehouses, with descriptions and tags.
Lineage A map of how tables flow through jobs and datasets. Groundzero records it for you when work runs.
AI assistant and agents The assistant answers questions and drafts work in chat. An agent is a saved assistant with one fixed job that other systems can call.

Before you start#

You need Why
An account on your organisation's Groundzero site Your administrator creates accounts. The sign-in page has no sign-up link.
An authenticator app (Google Authenticator or Authy) Groundzero asks you to set up two-step verification the first time you sign in.
Permission for what you want to do Your administrator grants access. When you lack it, Groundzero says so, for example You do not have "All" or "Add" permissions on Connection.
Details of one data source, for example a PostgreSQL server You connect it in step 1.

Not sure what you may do?

Try each step. Nothing in this guide changes the data in your sources, and when a permission is missing Groundzero names it. Send that message to your administrator.

How to use this guide#

Chapters 1 to 3 take about an hour and give you the first workflow from start to finish. Chapters 4 and 5 take about 20 and 15 minutes each, and you can do them whenever you are ready. Chapter 6 points you to what to learn next.

About the screenshots

The screenshots show sample data for a company called Acme Retail, so a name in one figure can differ slightly from the same step in another. Follow the instructions and the labels on your screen rather than matching names exactly.