Introduction

Margin is a data science workbench that keeps your notebooks, datasets, and narratives connected end‑to‑end—with an AI agent that works alongside you.

The Problem

Analysis breaks in the handoff. You do real work in a notebook, then copy charts into slides, paste tables into docs, and screenshot results for stakeholders. By the time your analysis reaches anyone else, the connection to the actual work is gone.

Numbers get copied. Charts get pasted. Assumptions get lost.

The Solution

Margin keeps everything in one place:

• Cloud notebooks – Real Python environments. No local setup. Open and start working.
• AI agent – An assistant that can write code, run cells, and iterate on analysis alongside you.
• Connected briefs – Charts, tables, and results stay linked to the exact code that produced them.
• Shareable outputs – One link. Interactive. No PDFs. No permission gymnastics.

Core Concepts

Notebooks

Notebooks are your private workspace for exploration. Write real Python code, import data, create visualizations, and iterate freely. Margin notebooks work like Jupyter—code cells, markdown cells, and cell outputs you're already familiar with.

Briefs

Briefs are stakeholder‑ready documents built from your notebook work. Instead of copying and pasting, you promote specific outputs (charts, tables, images) from your notebook directly into a brief. Every artifact stays traceable to its source.

Datasets

Upload CSV, JSON, JSONL, or Parquet files to Margin. Your datasets live in secure cloud storage and are accessible from any notebook in your workspace.

How It Works

Next Steps

• Create your account and set up your workspace
• Follow the quickstart to create your first notebook and brief
• Work with the AI agent to accelerate your analysis
• Learn about sharing to publish your work