Why ChartBook?

Why ChartBook?#

Data science teams face a common challenge: analytics work is scattered across notebooks, scripts, and shared drives. Documentation is an afterthought. When someone needs to find existing analysis or understand how a dataset was created, they spend hours searching or rebuild from scratch.

ChartBook solves this by providing a centralized platform for cataloging and documenting data science work.

The Backstage Analogy#

If you’re familiar with Backstage, Spotify’s open-source developer platform, ChartBook serves a similar purpose for data science teams.

Backstage (Software Teams)

ChartBook (Data Science Teams)

Software catalog

Pipeline catalog

Service documentation

Chart and dataframe documentation

API references

Dataset references

Tech docs

Analytics documentation

Templates and scaffolding

Pipeline templates

Where Backstage helps software engineers discover microservices and APIs, ChartBook helps data scientists discover pipelines, datasets, and analytics.

Key Benefits#

Discovery#

Find existing work before starting new analysis. chartbook builds searchable documentation sites where team members can browse pipelines, charts, and datasets.

Documentation#

Documentation is generated from your pipeline configuration, not maintained separately. Define your charts and dataframes in a TOML file, and ChartBook builds a complete documentation website.

Governance#

Track data sources, licenses, and access permissions for every dataset. Know where your data comes from, who can access it, and what restrictions apply.

Reproducibility#

Each pipeline is self-contained with its code, configuration, and outputs. Version control your pipelines and regenerate documentation as work evolves.

Programmatic Access#

Load any cataloged dataset directly into your analysis:

from chartbook import data

df = data.load(pipeline="sales", dataframe="quarterly_summary")

Who It’s For#

Data science teams building recurring analytics and reports.

Analytics engineers maintaining data pipelines and transformations.

Research teams producing charts and datasets that need to be discoverable.

Organizations that need governance and traceability for their analytics work.

Project Types#

ChartBook supports two project types:

  • Pipeline — A single analytics pipeline with its own charts, dataframes, and documentation (type = "pipeline")

  • Catalog — A collection of multiple pipelines aggregated into a unified documentation site (type = "catalog")

Start with a pipeline to document a single project, then create a catalog when you want to aggregate multiple pipelines into a searchable portal.

How It Works#

  1. Organize your work into pipelines — Each pipeline contains code, data, and outputs

  2. Configure with TOML — Define charts, dataframes, and metadata in chartbook.toml

  3. Build documentation — Run chartbook build to generate an HTML site

  4. Share and discover — Host the site internally for your team to browse

See the Getting Started guide to create your first pipeline.

Contents