Show HN: Datapane – Build data science documents using Python | Hacker News Reader

Show HN: Datapane – Build data science documents using Python

datapane.com

5 pointspea1 comment

Hello HN,

We're Leo and Mandeep from Datapane.

We've built a way to create interactive data science documents, such as reports, dashboards, and articles, using Python. Instead of copying screenshots into PowerPoint or rebuilding in a BI tool, you can import Datapane's framework into your notebook or script and generate shareable documents in a few lines of code.

We started hacking on Datapane at our previous start-up - we loved working in the Python data stack, but struggled to share results with our clients or team. Sending a notebook didn't work because they didn't want a code-heavy analysis or have the environment (dependencies, access to data, etc.). Instead, we manually created decks, which had a high fixed-cost and were hard to automate.

We built Datapane to allow us to create interactive documents from plots (using interactive libraries like Altair), DataFrames, and files. The library has come a long way since, and now supports components like drilldown tables, tabs, grid layout, drop downs (example: https://datapane.com/u/datapane/reports/sample-external-repo...)

Datapane is different to app frameworks like Dash, in that it bakes everything into a standalone HTML document. You can publish reports from existing environments (e.g. Colab, Airflow, or PySpark), and your reports don't need a running server - it's like an interactive PDF.

We launched a public platform so people could share reports and embed them into places like Medium. Since then, we've been humbled by the things our community have built, which you can see in our gallery: https://datapane.com/explore

We make money by charging companies who need to share reports privately (or with clients) or need to automate report generation using our Script Automation API. This lets you deploy your notebooks and scripts, so your team run them with parameters using web-forms or schedule them.