One Mindmap to Learn Anything

Reading research papers is a pain. We’re scrolling through pages, trying to connect ideas across sections, and by the time we finish, we’ve forgotten half of what we read. It’s like trying to build a house while someone keeps moving the foundation.

Traditional reader apps treat documents as static, linear text. But knowledge isn’t linear—it’s interconnected. When I was researching, I always found myself creating messy hand-drawn mind maps just to understand how concepts related to each other.

That’s when I realized: what if we could automatically extract the structure from any document and turn it into a mind map? Even better, what if the mind map is interactive and we can ‘ask’ follow up questions?

So I created Papermap, an app that does just that. You can try it out here:

apps.raihankalla.id/papermap

Here’s the launch video - Part 1:

Introducing Papermap —my 13th experimental app.

This is how students & researchers can read & understand any paper 10x faster.

Turn any PDF into an interactive AI mindmap, getting you up to speed on the entire paper at a glance, then dive into the details visually.

Link ↓ https://t.co/FRrvMUg456 pic.twitter.com/ehMqOsKt91

— Al Harkan (@alhrkn) April 6, 2025

Part 2:

What if ChatGPT always answered in mindmaps?

This is the next iteration of Papermap. It can now answer any question and process any URL.

Enjoy ↓ https://t.co/SkO6OwCZ0I pic.twitter.com/Fm1WxT7TZs

— Al Harkan (@alhrkn) April 15, 2025

And here’s the brief technical overview on how it works:

graph TB
    A[User Input] --> B{Input Type}
    B -->|PDF| C[PDF Upload & Storage]
    B -->|Text| D[Direct Text Processing]
    B -->|URL| E[Web Content Extraction]
    
    C --> F[Content Extraction]
    D --> G[AI Analysis]
    E --> H[Content Fetching]
    
    F --> G
    H --> G
    
    G --> I[Structured Mind Map]
    I --> J[React Flow Visualization]
    J --> K[Interactive Mind Map]
    
    L[Google Generative AI] --> G
    M[Vercel Blob Storage] --> C
    N[PostgreSQL + Drizzle] --> I

Most AI tools give you a summary or extract key points. Papermap creates a hierarchical structure that reflects how concepts actually relate to each other. It’s like having a research assistant who not only reads the paper but also creates a perfect outline.

And the best part, the mind map isn’t static. You can:

• Click nodes to expand descriptions
• Collapse branches to focus on specific areas
• Ask follow-up questions about any node
• Navigate between the mind map and original content
• Switch between different layout styles

You can also provide more than just PDFs. You can input text questions, paste URLs, or upload documents. This makes it versatile for different use cases—from academic research to content analysis to idea organization.

I hope this tool is useful for you. If you have any questions or feedback, please let me know @alhrkn.