How to Get Started with Deepnote for Data Analysis

What Deepnote Does and Why It Matters

Deepnote is a browser-based data notebook designed for teams that need to analyze data together. Think of it as Google Docs for data work: multiple people can edit the same notebook at the same time, leave comments on specific cells, and share results with stakeholders who never need to touch code.

The platform supports Python, SQL, and R in code blocks, but it also offers no-code cells for charts, pivot tables, and input parameters. This makes it useful for mixed teams where some members write queries and others just need to interact with the output.

Since its launch in 2019, Deepnote has grown to over 500,000 users. In 2026, the company open-sourced its notebook format under the Apache 2.0 license, which means you can run Deepnote locally in your IDE and optionally sync to the cloud for collaboration features.

Creating Your Account and First Project

Start at deepnote.com and sign up with a Google, GitHub, or email account. The free tier includes up to three editors, five projects, unlimited viewers, and basic AI code assistance.

Once inside, click "New Project" in the sidebar. Each project is a self-contained workspace with its own notebooks, data connections, and environment settings. Name it something descriptive like "Q1 Revenue Analysis" or "Customer Churn Review" so teammates can find it later.

Deepnote creates a default notebook inside each project. Notebooks are made up of blocks, and each block can be a different type: Python code, SQL query, Markdown text, a chart, or an input parameter. You can drag blocks to reorder them.

Connecting Your Data Sources

The real power of Deepnote shows up when you connect it to where your data lives. Click the "Integrations" tab in the left sidebar, then "Create new."

Deepnote supports direct connections to PostgreSQL, MySQL, BigQuery, Snowflake, Redshift, ClickHouse, MongoDB, and several other databases. For each connection, you provide the host, port, database name, and credentials. Once saved, the connection appears in your sidebar with a schema browser, so you can preview tables and columns before writing a single query.

If your data lives in files rather than a database, you can upload CSVs, Excel files, or Parquet files directly into a project. Deepnote stores them in the project's file system, and you can read them into a Python or SQL block immediately.

For teams using dbt, Deepnote also integrates with dbt Core and dbt Cloud. This lets you reference dbt models directly in your SQL blocks and keep your transformation logic consistent.

Writing Your First Analysis

With data connected, add a SQL block to your notebook and write a query against your database. Results appear as an interactive table below the block. You can sort columns, filter rows, and export to CSV without writing additional code.

To create a visualization, click "Add visualization" below any result table. Deepnote offers bar charts, line charts, scatter plots, histograms, and pivot tables as no-code blocks. Select your axes, group-by fields, and color schemes from dropdown menus. No Python or JavaScript required.

For more complex analysis, switch to a Python block. Deepnote comes with pandas, numpy, matplotlib, seaborn, scikit-learn, and plotly pre-installed. The AI assistant can generate code from a plain-English description if you prefer not to write Python by hand. On the free plan, you get 10 AI code completions per month. The Team plan at $39 per editor per month (billed annually) removes that limit.

A practical example: if you want to calculate monthly revenue trends from a sales table, you could write a SQL block to pull the raw data, then a Python block with pandas to resample by month and calculate growth rates, followed by a no-code chart block to visualize the trend. Each block passes its output to the next, creating a readable pipeline from raw data to insight.

Collaborating with Your Team

Invite teammates by going to "Settings and members" in the left sidebar. You can share a workspace-level invite link or add people by email. Deepnote supports three roles: Viewer (read-only), Editor (can modify notebooks), and Admin (manages settings and billing).

Real-time collaboration works like a shared document. You see teammates' cursors, and changes sync instantly. One important guardrail: two people cannot edit the same block at the same time, which prevents merge conflicts in code cells.

Comments work on a per-block basis. Click the comment icon next to any block to start a thread. This is useful for code reviews, flagging data quality issues, or asking a teammate to verify a number before it goes into a report.

Version history is available on all plans (7 days on Free, 30 days on Team). You can restore any previous version of a notebook, which acts as an undo mechanism for the entire project.

Sharing Results Beyond Your Team

Not everyone who needs your analysis should be inside Deepnote. The platform offers three ways to share:

Published apps let you turn a notebook into an interactive web page. Viewers can adjust input parameters (date ranges, filters, dropdown selections) and see the results update live, without seeing any code. This is useful for weekly reports that executives or clients review.

Scheduled runs let you set a notebook to execute on a recurring basis, such as every Monday morning. Combined with integrations like Slack or email, you can automate a reporting workflow where a notebook runs, generates updated charts, and sends the output to a channel.

Export options include PDF, HTML, and direct links. If you need a static snapshot for a board deck or a compliance record, PDF export captures every block in its current state.

Running Deepnote Locally

Since the open-source release, you can install Deepnote's notebook format locally and work in VS Code, JetBrains, or any editor that supports the extension. Run pip install deepnote to get started. Local notebooks use the .deepnote format but remain backward-compatible with .ipynb files, so you can convert existing Jupyter notebooks without losing content.

The local setup is useful for working offline, running on sensitive data that cannot leave your network, or prototyping before pushing work to the cloud. When you are ready to collaborate, push the notebook to Deepnote Cloud and your teammates can pick up where you left off.

Practical Tips for Getting the Most Out of Deepnote

Keep one notebook per analysis question. It is tempting to pile everything into a single file, but separate notebooks are easier to find, share, and schedule.

Use Markdown blocks to document your reasoning between code blocks. When a teammate opens your notebook six months later, the narrative context saves hours of reverse engineering.

Pin the data connection to the project level rather than embedding credentials in code. This keeps secrets out of notebooks and makes it easy to swap staging and production databases.

If you want to skip the setup and get straight to analysis from a file upload, tools like VSLZ handle the full pipeline from data ingestion to charts and statistical output in a single prompt, with no configuration required.

Summary

Deepnote provides a collaborative, browser-based environment for data analysis that bridges the gap between technical and non-technical team members. The free tier is generous enough for small teams to evaluate the platform fully. Start by connecting a data source, write a SQL query, add a no-code chart, and share the published app with a stakeholder. From there, layer in scheduling, AI assistance, and team workflows as your needs grow.