How I Built an Edge AI Pipeline to Automate Podcast Admin

Описание и применение

Пост‑обработка подкастов сократилась с двух часов до пятнадцати минут благодаря Edge‑AI‑конвейеру на Cloudflare Workers, использующему Llama‑3‑8b‑instruct для генерации заметок, заголовков и SEO‑ключевых слов.

Задача

Post‑recording administrative tasks consumed 80% of the creator’s time, causing burnout and inconsistent episode quality.

Результат

The pipeline reduced post‑production time from two hours to about fifteen minutes, improving consistency and allowing the creator to focus on content creation.

Рабочий процесс

• 1. Record episode
• 2. Transcribe audio
• 3. Send transcript to Cloudflare Worker for show‑notes generation via Llama‑3‑8b‑instruct
• 4. Receive structured JSON with summary, takeaways, timestamps, resources
• 5. Review and adjust timestamps
• 6. Generate episode title using prompt‑engineered title generator
• 7. Retrieve SEO keyword suggestions
• 8. Publish episode metadata

Доказательства

I would spend two hours writing show notes, agonizing over the episode title, and trying to extract SEO keywords from the conversation.

Источник

The results have been a massive improvement for my own workflow. What used to take me two hours of post‑recording admin now takes about 15 minutes.

Источник

By explicitly banning clickbait that overpromises and enforcing a word count, the output quality jumped significantly.

Источник