I Spent Hours Writing Show Notes, So I Built an Edge AI Pipeline

Description and use

Automating podcast metadata creation with an edge‑AI pipeline cuts post‑production time and improves quality.

Problem

Podcast creators spend hours converting raw transcripts into show notes, titles, and descriptions, causing cognitive overload and operational burnout.

Outcome

The pipeline reduces post‑production time from hours to minutes, improves metadata quality, and enhances podcast discoverability.

Workflow

• 1. Capture raw transcript from recording
• 2. Send transcript to Cloudflare Workers AI endpoint
• 3. AI model (Llama 3 8B Instruct) processes transcript using a system prompt that defines markdown structure for show notes
• 4. AI returns JSON with episode summary, key takeaways, timestamps, and resources
• 5. Separate title generator uses a prompt that enforces hooks, word limits, and tone
• 6. Optional tools generate episode outline, description, and SEO keywords
• 7. Output is displayed in UI for quick review and publishing

Evidence

I was struggling heavily with the post‑production admin. Specifically, I’d spend hours staring at a massive, unformatted 45‑minute transcript, trying to extract the "good parts" for show notes, craft a catchy title, and write an SEO‑friendly description.

Source

I decided to build PodCrisp using Cloudflare Workers, leveraging their edge AI inference. Running AI at the edge means the heavy lifting happens close to the user.

Source

The magic here is in the system prompt. By explicitly defining the markdown structure, the model knows exactly how to format the output.

Source

Moving to an edge-native architecture paid off immediately. The UI feels incredibly responsive because the inference happens at the edge, and the structured show notes are surprisingly accurate right out of the box.

Source