Fabian Westhäußer

Ruby News | AI News Reporter

The AI-supported music-news workflow behind Ruby News turns traceable sources into long-form and short-form content for five platforms.

Visit project
For
Ruby News (public media brand)
Project type
AI-supported music-news workflow
Industry
Media / music news
Scope
Research, story filtering, , long-form composition, AI-supported short ranking, voice generation, video creation, attribution, and multi-platform publishing
Visibility
Public brand and content; operator private
Timeline
In production since November 2025
Conceptual view of the production logic: the long-form rundown is composed around pacing and topic variety and published on YouTube and Spotify. Every short goes to YouTube; the three highest-ranked also go to TikTok, Instagram, and Facebook.

The challenge

Ruby News is a public music news channel on YouTube, with distribution across Spotify, TikTok, Instagram, and Facebook. The AI News Reporter is the workflow behind it, from source to publication, turning traceable sources into scripts, videos, and platform-specific publications.

The project started from scratch. The challenge was to build a complete that could turn current music news into publishable video content in a structured and repeatable way. That meant handling research, topic selection, , visual assembly, platform-specific output, and publishing in one connected system.

The solution

I designed and built an AI-supported that collects music news, groups related reports, documents facts and sources, and creates scripts, voice, and video assets. For long-form episodes, the system composes a balanced rundown around pacing, topic variety, and transitions. Short stories are ranked separately with AI support based on their expected audience interest: every short is published on YouTube, while the three highest-ranked also go to TikTok, Instagram, and Facebook. Long-form content is published on YouTube and Spotify.

Result

Since 11/2025in production
for long-form and short-form output

Since November 2025, the workflow has continuously produced source-based music news for Ruby News. It combines research, two distinct editorial decision models, media production, attribution, and long-form and short-form distribution across five platforms.

Implementation

My role

  • concept
  • development
  • layout design
  • maintenance

From article to publication

Sources & editorial work

  • automated collection of music-related news articles
  • story grouping and tracking across related articles
  • fact extraction and script generation
  • attribution and source documentation

Format & selection

  • AI-supported ranking of each story by expected audience interest
  • composition of balanced long-form rundowns around pacing, topic variety, and transitions
  • generation of shorter story-based video versions

Media production

  • AI-supported voice generation and transcript timing
  • subtitle creation based on word-level timestamps
  • media cue generation and stock media sourcing
  • video assembly with avatar overlay, titles, subtitles, and background audio
  • thumbnail generation

Publication

  • publication of every short on YouTube
  • automatic selection of the three highest-ranked shorts for TikTok, Instagram, and Facebook
  • long-form distribution on YouTube and Spotify

Tech

Python NewsAPI Tavily Azure OpenAI embeddings ElevenLabs ffmpeg OpenCV Pillow Google image generation tools YouTube API Meta

Relevant services

design custom software development editorial video content development multi-platform publishing maintenance and iteration of internal production workflows

Project views

The public record for the 18 May 2026 edition connects the news sources used with the corresponding media attributions.
The same long-form edition was published with its original Ruby News artwork on YouTube and Spotify .
One short, four authentic publications: YouTube , TikTok , Instagram , and Facebook . Background image: “Close-up of a Billboard Award” by RDNE Stock project via Pexels .

If you need a custom AI-supported for turning ongoing information into structured content and publishable media, this project shows the kind of workflow I can build.

Discuss a project
Contact