← Back to portfolio Content Automation

Automated News Repurposing & Content Re-Digestion for a Digital Publisher

Automated content engine for a digital news website. The system continuously monitors trending topics across news APIs, social signals, and the publication's own audience analytics, then mines the publisher's back catalogue and connected YouTube channels for relevant evergreen material. Long-form articles and video segments are transcribed (Whisper), semantically re-organised, and re-drafted by an LLM into fresh angles tied to the current trend — with full source citation and editorial guardrails. Drafts land in the editorial CMS for journalist review and one-click publish, restoring the value of work that would otherwise sit in the archive.

Outcome

Publishing throughput rose substantially without growing the editorial team. Older content (articles + YouTube transcripts) now contributes meaningfully to current traffic instead of sitting in the archive.

Tech stack

PythonFastAPIOpenAI WhisperAnthropic ClaudeYouTube Data APIPostgreSQLpgvector

Discuss a similar project

30 minutes, no pitch.

Book a Free Call →