- The two requirements for the first tutorial
- The quick-start tutorial, one move at a time
- How a Databricks workspace is organised
- What the getting-started docs cover
- Genie One for business users
- Beyond the browser: the API and CLI
- Where to go after your first chart
- A first-session checklist
- Frequently asked questions
To use Databricks, sign in to a workspace, click New then Notebook, query the sample table samples.nyctaxi.trips and press Shift+Enter to run it. From there you can turn the results into a chart, and the vendor's other tutorials take you into data engineering and machine learning.
The vendor describes its get-started tutorials as an introduction to core features that guides you through the basics of working with the Databricks platform. The first one is small: one notebook, one sample table, one chart. This page follows that route.
Everything below comes from the vendor's own documentation, the AWS version in particular. The docs are split by cloud, so if you're on Azure or Google Cloud the steps may differ. Keep the docs for your cloud open alongside.
The two requirements for the first tutorial
The query tutorial lists two requirements. Your workspace needs Unity Catalog enabled, and you need permission to use an existing compute resource or create a new one. If you don't have that permission, the docs say to ask your Databricks administrator.
You'll also need a Databricks account before any of the tutorials will work. The docs point to signing up if you don't have one yet.
The quick-start tutorial, one move at a time
- Sign in to your Databricks workspace.
- Confirm Unity Catalog is enabled and you can use or create compute.
- Click New in the sidebar, then Notebook. A blank notebook opens.
- Write a query against the sample table
samples.nyctaxi.tripsand press Shift+Enter to run the cell. - Next to the Table tab, click the plus, then Visualization, and pick Bar.
- Choose the X column, the Y column, an aggregation and a group-by column, then click Save.
- Pick your next tutorial: importing CSV data, building an ETL pipeline or trying the AI Playground.
At that point you've created a notebook, queried a sample table and saved a chart, which is everything the first tutorial covers.
How a Databricks workspace is organised
The concepts page explains the terms the rest of the docs use. It describes a workspace as a cloud deployment that acts as an environment for your team to access Databricks assets, and it says one account can include multiple workspaces.
Data sits inside Unity Catalog, which the concepts page calls the governance layer. It works as a hierarchy: catalogs at the top, schemas inside catalogs, and tables, views and volumes inside schemas. Accounts with Unity Catalog enabled can manage users and their data access centrally across every workspace in the account.
The concepts page also says that, by default, all tables created in Databricks are Delta tables.
What the getting-started docs cover
| Area | What the vendor's docs say |
|---|---|
| Tutorial groups | Data exploration, data engineering, and AI and machine learning |
| Data exploration | Query and chart data in a notebook, import CSV data, create a table, try Genie One |
| Data engineering | Build an ETL pipeline with Lakeflow pipelines and Auto Loader, or with Apache Spark |
| AI and machine learning | Train a scikit-learn model tracked with MLflow; query large language models in the AI Playground |
| Workspace | A cloud deployment your team uses to access Databricks assets; one account can hold several |
| Unity Catalog | Catalogs, then schemas, then tables, views and volumes |
| Default table type | Delta tables |
| Ways in | Web UI, REST API, SQL REST API and the Databricks CLI |
Genie One for business users
If you're a business user who wants answers rather than notebooks, the docs describe Genie One as a simplified interface. It offers AI/BI dashboards, data questions and Databricks Apps without the technical workspace concepts above. There's a tutorial for it in the data exploration group.
The quick-start page also offers a Genie Code prompt that builds the notebook for you. The steps on this page are the manual version of the same tutorial.
Beyond the browser: the API and CLI
The web UI isn't the only way in. The concepts page lists a REST API, a SQL REST API and the Databricks CLI alongside it.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
The concepts page also says usage is measured in Databricks units (DBUs), and that account-level administration covers billing and support. If you're not the account admin, that side of things sits with whoever is.
Where to go after your first chart
The remaining tutorials are grouped by type of work. Data exploration covers importing and visualizing CSV data, creating a table and using Genie One. Data engineering covers building an ETL pipeline with Lakeflow pipelines and Auto Loader, or with Apache Spark. AI and machine learning covers training a scikit-learn model tracked with MLflow, and the AI Playground, where the docs say you can query large language models and prototype a tool-calling agent with no code.
If you're still deciding whether Databricks is the right platform, my comparison of Databricks vs Snowflake looks at how the two differ, and this list of Databricks alternatives covers other lakehouse platforms. For lighter work, these data analytics tools may be all you need.
A first-session checklist
Get an account, confirm Unity Catalog is enabled and that you can use or create compute, then open a notebook, query the taxi sample and save a bar chart. After that, the concepts page explains catalogs, schemas and workspaces, and the tutorial groups above point you to the next step for your kind of work.
Frequently asked questions
Do I need to write code to use Databricks?
The quick-start tutorial has you write a query against a sample table. For business users, the docs describe Genie One as a simplified interface without the technical workspace concepts.
What do I need before the first tutorial?
A Databricks account, a workspace with Unity Catalog enabled and permission to use or create compute. If you're missing permission, the docs say to ask your Databricks administrator.
How do I create a notebook in Databricks?
Click New in the sidebar, then Notebook. A blank notebook opens, ready for your first query.
What is a Databricks workspace?
The concepts page calls it a cloud deployment that works as an environment for your team to access Databricks assets. One account can include several workspaces.
What type of table does Databricks create?
By default, every table created in Databricks is a Delta table, according to the vendor's concepts page.
