YouTube Intelligent Scraper
A system that scrapes YouTube channels, scores relevance with AI, and suggests a watch order.
[ The transformation ]
Working through a large channel’s back catalogue is guesswork — you can’t tell which videos are worth your time first.
A channel’s back catalogue becomes a ranked, relevance-scored watch list instead of a wall of thumbnails.
[ The hard problem ]
Scoring hundreds of videos for personal relevance using AI — pulling metadata and transcripts, running each through a relevance classifier, and presenting a ranked watch order in real time.
[ How it works ]
I built a Node.js pipeline that pulls every video + metadata from channels, downloads transcripts, AI-scores each video against your interests, prioritises a watch order, and serves it in a filterable dashboard.
[ Key features ]
Full channel scraping — every video, metadata and transcript
AI relevance scoring against user interests
Prioritised watch order based on scores
Filterable dashboard for browsing results
[ The result ]
