How to Extract Tools and Technologies from YouTube Videos Automatically

Description and use

An automated pipeline extracts tool names, categories, quotes, timestamps, and official links from YouTube videos, improving recall over large chunks and avoiding incorrect links.

Problem

Technical YouTube videos mention many tools, commands, and versions, but viewers cannot recall them after the video ends.

Outcome

The pipeline reliably extracts tool names, categories, quotes, timestamps, and official links from YouTube videos, improving recall over large chunks and avoiding incorrect links.

Workflow

• Get the transcript using yt-dlp and keep timing data.
• Split the transcript into manageable chunks (≈30,000 characters) to avoid model summarisation and silent truncation.
• Extract candidate tools with structured output, requiring name, category, quote, and timestamp, and deduplicate after merging chunks.
• Verify each candidate by searching with category and description, accept canonical repos, block aggregator sites, score matches, and return no link if confidence is low.

Evidence

Most videos have captions, either written by the author or generated automatically. Libraries such as yt-dlp can fetch them along with timing data.

Source

We found chunk size matters more than expected. In our tests on a transcript of roughly 475,000 characters, small chunks found more tools than large ones.

Source

Ask the model for structured output, not prose. For each candidate tool, request the name, a short description, a category and a verbatim quote from the transcript with its timestamp.

Source

What worked for us: Build the search query from the tool's category and description, not only its name.

Source