Building an AI Podcast Workflow: My Retrospective and Mistakes

Описание и применение

Автоматизация пост‑продакшн для подкастов с помощью Edge‑AI на Cloudflare Workers снижает административную нагрузку и повышает SEO‑видимость.

Задача

Podcast post‑production is time‑consuming and hampers discoverability, causing creator burnout.

Результат

The edge AI workflow cuts post‑production time, improves title relevance and show‑note structure, and enhances discoverability, thereby reducing creator burnout.

Рабочий процесс

• 1. Record episode and export audio
• 2. Generate transcript (outside scope)
• 3. Send transcript to Cloudflare Worker
• 4. Worker calls @cf/meta/llama-3-8b-instruct with structured system prompt
• 5. Receive JSON with formatted show notes
• 6. Use separate title generator function to produce numbered title list
• 7. Publish episode with generated title and show notes

Доказательства

I was spending 80% of my time on administrative overhead and only 20% on actual creative content.

Источник

I decided to build the entire platform on Cloudflare Workers, utilizing AI inference at the edge.

Источник

The latency is incredibly low, meaning users get their generated titles and show notes almost instantaneously, regardless of where they are in the world.

Источник