Project Plan: From Markdown to Interactive Charts | Generated by AI

Home 2026.06

Nice! So the plan is clear:

When you’re ready to start the HTML conversion, a few things worth thinking through:

Architecture options:

  1. Static site generator (Jekyll/Hugo) → output HTML with embedded JS charts — least friction since you’re already on GitHub Pages
  2. Build a small Python script that takes your markdown posts + AI summary → generates rich HTML with D3/Chart.js visualizations per post
  3. Hybrid — keep markdown as source of truth, add a render layer on top

Given your “ship fast” style, option 2 fits best — a md2html.py that calls Claude API to extract key data points from each post and auto-generates charts. You could batch-process your 400 posts overnight on the workstation.

What makes blog posts actually visual:

When the model finishes training, share the results — curious what dataset and objective you used for the code model.


Back Donate