How to Get Started with Briefer for Data Analysis

Briefer is an open-source analytics notebook that gives teams SQL, Python, and dashboard tools inside a single Notion-like interface. It connects directly to PostgreSQL, BigQuery, Snowflake, Redshift, Athena, MySQL, and CSV files. The cloud version is free for up to three users and requires no local installation. The self-hosted version runs via Docker with a single command. Either way, you can go from sign-up to a working SQL query in under ten minutes.

Briefer launched as a Y Combinator S23 company and went open-source in mid-2024. Its GitHub repository crossed 14,000 stars within a year of going open-source, making it one of the fastest-growing analytics tools on GitHub in 2024. The platform's AI analyst agent, launched in late 2024, can write SQL queries, generate charts, and run Python code from plain-language prompts without requiring any coding knowledge from the user.

Who Briefer Is For

Briefer is designed for teams where analysts and non-technical stakeholders work from the same data. Data analysts write the SQL. Operations managers, marketers, or founders can filter and explore those results through interactive input controls built directly into the notebook without touching the underlying code. This model reduces back-and-forth on data requests and keeps all reports in a single shared workspace.

The audience most underserved by Briefer's existing documentation is the non-technical side: operations managers, marketing analysts, and founders who need data answers quickly without configuring a full development environment. This guide covers the fastest path for that audience.

Option 1: Start on Briefer Cloud

The fastest way to start is the cloud version at briefer.cloud. Sign up with an email address, create a workspace, and invite teammates. The free tier supports up to three users with unlimited notebooks and dashboards.

Once inside, Briefer opens to a notebook editor. Click "New notebook" to create your first one.

Connect a data source: In the sidebar, click Data Sources, then "Add data source." Briefer lists PostgreSQL, BigQuery, Snowflake, Redshift, MySQL, Athena, Trino, and SQL Server. Select your database type, fill in the host, port, database name, user, and password, then click Test Connection. Briefer runs a connectivity check and shows a confirmation when successful.

If you do not have a database, upload a CSV or Excel file directly from the notebook. Click the "+" button inside the notebook, select "File upload," and drag your file in. Briefer stores it in an internal DuckDB instance and makes it immediately queryable with standard SQL.

Option 2: Self-Host with Docker

For teams that cannot send data to third-party cloud services, Briefer runs entirely on your own infrastructure. The only requirement is Docker.

Run this single command in your terminal:

docker run -d \
  -p 3000:3000 \
  -v briefer_psql_data:/var/lib/postgresql/data \
  -v briefer_jupyter_data:/home/jupyteruser \
  -v briefer_briefer_data:/home/briefer \
  briefercloud/briefer

Briefer will be available at http://localhost:3000. The Docker image includes its own PostgreSQL database for internal state and a Jupyter kernel for Python execution. You do not need to install Python, Jupyter, or any data science libraries separately.

Alternatively, install via pip:

pip install briefer
briefer

This command pulls the Docker image and starts the application automatically. Data is stored in Docker volumes and persists across restarts.

Running Your First SQL Query

Open a notebook and click the "+" button to add a block. Select "Query." Briefer shows a SQL editor with autocomplete for your connected data source's tables and columns.

A typical first query to check recent activity:

SELECT
  date_trunc('week', created_at) AS week,
  COUNT(*) AS new_signups
FROM users
WHERE created_at >= NOW() - INTERVAL '90 days'
GROUP BY 1
ORDER BY 1;

Press Cmd+Enter (Mac) or Ctrl+Enter (Windows) to run it. Results appear as a table directly below the block. Briefer renders up to 10,000 rows in the browser without performance issues on most modern hardware.

For CSV uploads, the query syntax is the same standard SQL. Briefer exposes each uploaded file as a table named after the filename. A file called orders.csv becomes the table orders.

Creating a Visualization

Below the query result, click "Add visualization." Briefer opens a point-and-click chart builder. Set the X axis to "week" and the Y axis to "new_signups." Select bar or line from the chart type dropdown. The chart updates live as you adjust the settings.

No code is required at this step. The chart is linked to the query above, so re-running the query with different parameters automatically refreshes the visualization.

For more complex charts, add a Python block below the query. Briefer automatically converts the SQL result into a pandas DataFrame using the query block's name as the variable. A block named "weekly_signups" becomes the DataFrame weekly_signups in subsequent Python blocks, letting you use matplotlib, plotly, or altair without any manual data loading.

Using the AI Analyst Agent

Briefer's AI agent is accessible from any notebook by clicking the agent icon in the top bar. Type a question in plain English, for example: "Which product categories drove the most revenue last quarter?"

The agent reads your connected schema and notebook context, writes the SQL, runs it, and generates a labeled chart with a summary paragraph. You can ask follow-up questions in the same thread. The agent handles chained analysis across multiple steps, including groupby, filtering, and statistical summaries.

The agent works within Briefer's execution environment and does not send your database schema or results to an external service by default. Teams on the self-hosted enterprise plan can configure the agent to use their own AWS Bedrock or Google Vertex AI credentials, keeping all computation inside their own cloud account.

In documented testing on a mid-size e-commerce dataset with 12 tables and approximately 800,000 rows, Briefer's agent correctly answered 78 percent of plain-language analytical questions on the first attempt without any manual SQL correction.

Sharing and Scheduling Reports

Notebooks convert to dashboards with one click. Toggle from "Notebook" mode to "Dashboard" mode in the top bar. In dashboard mode, code blocks are hidden and only visualizations, text blocks, and interactive input controls are visible to viewers.

To share externally, click "Share" and enable the "Public link" option. Anyone with the link can view the dashboard without a Briefer account. For internal sharing, invite teammates with Viewer or Editor permissions.

To schedule automatic refreshes, click "Schedule" in the top bar. Set an interval (hourly, daily, weekly) and optionally configure an email or Slack notification. Briefer re-runs the notebook on the defined schedule and sends a snapshot to the configured recipients.

Practical Tips for Non-Technical Teams

A few practices that make shared notebooks more useful for the full team:

Add text blocks between queries to explain the date range, filter logic, and any data assumptions. Non-technical readers rely on these annotations to interpret charts correctly without consulting the analyst who built them.

Use input blocks to let stakeholders filter by date range, region, or product themselves without needing to ask for a revised query. Input blocks are linked to query parameters and update the charts automatically when a filter changes.

Name every query block clearly. Briefer uses block names as Python DataFrame variable names. A block named "monthly_revenue" becomes the variable monthly_revenue in Python blocks downstream.

What Briefer Does Not Handle

Briefer is not a data pipeline or ETL tool. It queries existing databases and does not move data between sources or manage incremental syncs. If your source data sits in spreadsheets or email exports, you will need to load it into a database or upload it as a CSV file before Briefer can query it. For teams who want to skip the database setup entirely and go directly from a file upload to charts and statistical analysis, VSLZ handles that from a single prompt with no configuration.

Summary

Briefer gives teams a shared workspace for SQL, Python, and dashboards without the overhead of configuring a Jupyter server or a separate BI tool. The cloud version starts in under five minutes at no cost. The self-hosted Docker version adds data residency for teams in regulated industries. The AI agent reduces time to insight on routine analytical questions. For any team currently exporting database results to spreadsheets and building charts manually, Briefer is the most practical starting point available in 2026.