How to Set Up Quadratic for AI Data Analysis

What Quadratic Does Differently

Most spreadsheet tools force a choice: stay in formulas and hit a ceiling, or switch to a notebook environment and lose the grid. Quadratic removes that tradeoff. It is a browser-based spreadsheet where any cell can hold a Python script, a SQL query, or a JavaScript function alongside standard formulas. The output lands back in the grid, so you keep the familiar rows-and-columns view while running code that would normally require Jupyter or VS Code.

The platform runs entirely in the browser. There is no local install, no dependency management, and no environment configuration. Python cells execute on Quadratic's servers with common libraries like pandas, NumPy, and scikit-learn pre-installed. SQL cells connect directly to databases, and an AI assistant writes or fixes code when you describe what you need in plain English.

Step 1: Create Your Account

Go to app.quadratichq.com and sign up with a Google account or email. No credit card is required. After sign-in, you land in the file manager where you can create blank sheets or open templates.

Click "New Sheet" to open a fresh spreadsheet. The interface looks like a standard grid with one addition: a command palette accessible by pressing the forward slash key (/) in any cell.

Step 2: Import Your Data

Quadratic supports several data import methods depending on your source.

For CSV or Excel files, drag the file directly onto the sheet. Quadratic parses the file and drops the data into the grid starting at the selected cell. Column headers are preserved, and data types are detected automatically.

For database connections, click the "Connections" panel on the left sidebar. Quadratic supports PostgreSQL, MySQL, Microsoft SQL Server, and Snowflake. Enter your connection credentials, and the schema browser shows available tables and columns. You can then write SQL cells that query your database and return results to the grid.

For API data, use a Python or JavaScript cell to fetch from an endpoint. For example, a Python cell with a requests call can pull JSON data and return it as a DataFrame that populates the sheet.

Step 3: Write Your First Python Cell

Select any empty cell and press / to open the command palette. Choose "Python" from the list. A code editor panel opens on the right side of the screen.

Write a simple script to explore your data. If you imported a CSV into cells A1 through D100, you can reference that range:

# Reference spreadsheet data as a DataFrame
df = cells((0, 0), (3, 99))
df.describe()

The last expression in the cell is returned to the spreadsheet. In this case, df.describe() outputs summary statistics directly into the grid. Single values return to one cell; DataFrames expand across multiple cells.

Python cells have access to pandas, NumPy, SciPy, scikit-learn, and other common libraries without any import errors. You do not need to pip install anything.

Step 4: Use the AI Assistant

Click the sparkle icon in the toolbar or press Ctrl+K (Cmd+K on Mac) to open the AI Chat panel. This is where Quadratic's built-in AI becomes useful.

Type a request in plain English. For example: "Create a bar chart showing total sales by region from the data in columns A through D." The AI generates a Python cell with the appropriate pandas and matplotlib code, places it in your sheet, and runs it. The chart renders inline.

The AI assistant handles several types of requests well: data cleaning (removing duplicates, filling missing values, standardizing formats), statistical analysis (correlations, regressions, hypothesis tests), visualization (charts, heatmaps, distribution plots), and data transformation (pivots, merges, group-by aggregations).

If the generated code throws an error, you can ask the AI to fix it. Send a follow-up message like "that gave a KeyError on the Region column" and the AI revises the code with the correction.

Step 5: Run SQL Queries in the Sheet

If you connected a database in Step 2, you can write SQL cells that query it directly. Press / in a cell, select "SQL," and choose your connection from the dropdown.

Write a standard query:

SELECT region, SUM(revenue) as total_revenue
FROM sales
WHERE sale_date >= '2026-01-01'
GROUP BY region
ORDER BY total_revenue DESC

Results populate the grid starting from the cell where you placed the query. The data updates each time you re-run the cell, creating a live connection between your spreadsheet and your database.

You can then reference SQL output in Python cells for further analysis, chaining SQL data retrieval with Python-based processing in a single sheet.

Step 6: Combine Formulas, Code, and AI

The real strength of Quadratic is mixing all three modes in one workspace. A practical workflow looks like this:

First, use a SQL cell to pull raw transaction data from your database. Second, write a Python cell that cleans and aggregates the data. Third, use the AI assistant to generate a visualization. Fourth, add spreadsheet formulas in adjacent cells for quick calculations like running totals or percentage changes.

Each cell type can reference the output of any other. A formula cell can point to a value produced by a Python cell, and a Python cell can read data from SQL output. This makes Quadratic particularly effective for workflows that currently span multiple tools.

Step 7: Share and Collaborate

Click the "Share" button in the top right corner to generate a link or invite collaborators by email. Shared sheets support real-time multiplayer editing, similar to Google Sheets.

You can also export your sheet as a CSV for use in other tools, or keep it in Quadratic as a living document that updates whenever you re-run code cells or refresh database connections.

When Quadratic Works Best

Quadratic fits well for analysts who frequently hit the limits of spreadsheet formulas and switch to Python notebooks, teams that need to combine database queries with spreadsheet-style analysis in one view, and operators who want AI-assisted analysis without setting up a local coding environment.

It is less suited for large-scale production data pipelines or situations where you need full control over Python package versions. For those cases, a dedicated notebook or orchestration tool is a better fit.

If you want to skip environment setup entirely and go from a file upload to charts and insights from a single prompt, VSLZ handles that workflow with no configuration required.

Summary

Setting up Quadratic takes under five minutes: create an account, import data, and start writing Python, SQL, or formulas in any cell. The AI assistant fills the gap when you know what analysis you need but do not want to write the code from scratch. For teams that live in spreadsheets but need more power, it is a practical middle ground between a basic grid and a full data engineering stack.