← Back to list

Introducing Docling Studio: See What Docling Sees

I’ve been using Docling since the early days. If you work with document AI, you probably know how powerful it is — PDF parsing, structure…

Pier-Jean Malandrino in Scub-Lab · 2026-03-31 08:43 · 111 claps · 1.7 min read
#ai #programming #software-development #docling #ocr
Open on Medium ↗
Wiki topics: AI · AI · General 💻 · Programming

Introducing Docling Studio: See What Docling Sees

I’ve been using Docling since the early days. If you work with document AI, you probably know how powerful it is — PDF parsing, structure extraction, table detection, the engine handles it all.

But there was always one friction point: you couldn’t see what Docling was actually doing. Every time I needed to validate an extraction, I had to dig through JSON outputs, cross-reference bounding box coordinates, and mentally reconstruct the layout. It worked, but it felt like debugging with your eyes closed.

So I built Docling Studio.

What it does

Docling Studio is an open-source visual exploration and inspection layer for Docling outputs. The workflow is simple:

  • Upload a PDF
  • Configure your Docling pipeline (OCR, table extraction, enrichment options)
  • Run the extraction
  • Visually inspect every detected element — bounding boxes overlaid on the original pages, content preview on click, structured view of each extraction decision

The idea: make Docling’s outputs as readable as the documents it processes.

The stack

Vue 3 (Composition API) on the frontend, FastAPI on the backend, fully Dockerized for easy deployment. The repo has 180+ tests and CI/CD in place. It’s still early, but already usable.

Why I built it

Docling is an incredible piece of engineering.

I just wanted to build something that helps you realize how much it actually extracts. When you see the colored bounding boxes appear on your document, with each element typed and classified, you understand the depth of what the pipeline produces.

Try it

The project is fully open source: 👉 github.com/scub-france/Docling-Studio

Feedback, issues, and contributions are welcome. If you work with Docling or document AI in general, I’d love to hear what you think.

I am the CTO and Head of an architectural unit at Scub. I participate in the development of technological strategy, design solutions, and lead R&D projects.

Thank you for reading! If you enjoyed this article, please feel free to 👏 and help others find it. Please do not hesitate to share your thoughts in the comments section below.

Technology


메타데이터
post_id
bb43fbcefa8b
slug
introducing-docling-studio-see-what-docling-sees-bb43fbcefa8b
url
https://lab.scub.net/introducing-docling-studio-see-what-docling-sees-bb43fbcefa8b
canonical_url
https://lab.scub.net/introducing-docling-studio-see-what-docling-sees-bb43fbcefa8b
author_url
https://medium.com/@pmalandrino
status
ok
fetched_at
2026-06-10 12:26:30