The Problem
Organizations with large YouTube video libraries (training courses, webinars, product demos) cannot search within their own video content. YouTube's built-in search only matches titles and descriptions, not what was actually said in the video. Employees and customers searching for a specific topic discussed in a 45-minute webinar have to scrub through the entire video manually. The knowledge inside video content is effectively locked away.
The Scavio Solution
Build a searchable video knowledge base by using Scavio's YouTube endpoint to extract video metadata and then indexing transcripts into a search layer. For each video in your library, fetch metadata via Scavio, retrieve the auto-generated transcript, and index it into a vector store or full-text search engine. Users can then search for concepts and find the exact video and timestamp.
Before
Before: A training team had 200 YouTube videos. Employees searching for specific topics had to ask on Slack which video covered what. Average time to find the right video segment: 25 minutes. Knowledge reuse was low because finding information was harder than asking a colleague.
After
After: A searchable transcript index covers all 200 videos. Employees find the exact video and timestamp in under 30 seconds. Initial indexing cost: $1 (200 queries at $0.005). Weekly refresh for new videos: $0.10. Knowledge reuse increased measurably as measured by reduced repeat Slack questions.
Who It Is For
Training teams, knowledge management leads, and content librarians with large YouTube video libraries. Anyone whose team asks 'which video covered X?' on Slack multiple times per week.
Key Benefits
- Index 200 videos for $1 using YouTube metadata search
- Find exact video segments in seconds instead of 25-minute manual searches
- Weekly refresh for new content costs pennies
- Full-text transcript search catches topics mentioned but not in title/description
- Unlock knowledge trapped inside video content for the entire organization
Python Example
import requests
API_KEY = "your_scavio_api_key"
def search_video_library(query: str) -> list:
"""Search YouTube for videos matching query, return structured results."""
r = requests.post(
"https://api.scavio.dev/api/v1/search",
headers={"x-api-key": API_KEY},
json={"platform": "youtube", "query": query},
timeout=10,
)
data = r.json()
return [{
"title": v.get("title"),
"channel": v.get("channel"),
"link": v.get("link"),
"description": v.get("snippet", "")[:200],
"duration": v.get("duration"),
} for v in data.get("organic", [])[:10]]
# Search your channel's content
results = search_video_library("site:youtube.com/@yourchannel onboarding workflow")
for v in results:
print(f"{v["title"]} ({v["duration"]})\n {v["link"]}")JavaScript Example
const API_KEY = "your_scavio_api_key";
async function searchVideoLibrary(query) {
const res = await fetch("https://api.scavio.dev/api/v1/search", {
method: "POST",
headers: { "x-api-key": API_KEY, "content-type": "application/json" },
body: JSON.stringify({ platform: "youtube", query }),
});
const data = await res.json();
return (data.organic || []).slice(0, 10).map(v => ({
title: v.title,
channel: v.channel,
link: v.link,
description: (v.snippet || "").slice(0, 200),
duration: v.duration,
}));
}
const results = await searchVideoLibrary("site:youtube.com/@yourchannel onboarding workflow");
for (const v of results) {
console.log(`${v.title} (${v.duration})\n ${v.link}`);
}Platforms Used
YouTube
Video search with transcripts and metadata