How I Built an Edge AI Pipeline to Automate Podcast Notes

Description and use

An edge AI pipeline on Cloudflare Workers automatically generates structured show notes and titles for podcasts, cutting manual effort from hours to minutes.

Problem

Creating episode titles and show notes for a podcast was time‑consuming and mentally exhausting, creating a bottleneck after recording.

Outcome

The pipeline delivers instant, structured show notes and title options, cutting manual writing time from hours to minutes and improving SEO potential.

Workflow

• 1. Record podcast and generate transcript
• 2. Send transcript to Cloudflare Worker via JSON request
• 3. Worker calls Cloudflare AI endpoint with system prompt defining markdown structure
• 4. AI returns structured show notes (summary, bullet points, timestamps, resources)
• 5. Worker returns JSON to client
• 6. Separate title generator uses similar prompt for title options

Evidence

I was wrong. The real bottleneck happened long after I hit "stop recording."

Source

I decided to build PodCrisp using Cloudflare Workers.

Source

We use a system prompt to set strict rules. Instead of just saying "write notes," we explicitly define the markdown structure we want: a short summary, bullet points for takeaways, estimated timestamps, and resources.

Source

Because the inference happens on Cloudflare's network, the tool feels instantaneous to the user.

Source