YouTube Transcript Agent
Scrape, process, and chat with YouTube videos like never before.
O.O
Anonymous
social-media
0 downloads
Getting started with YouTube Transcript Agent
Step-by-Step Instructions
1
Airtable Setup
Create an Airtable Base Table name: YouTube Videos Fields: Video URL (single line text) recordId (autofilled) Transcript (long text or rich text) Status (e.g., “pending,” “processed”) (Optional) Cleaned Transcript, Vector Uploaded Add Button Field Add a button field to trigger the transcript scraping via script (using Run Script action).
2
Script Automation from Airtable
Add Script Block Paste your provided Airtable script into the scripting block. Replace: let webhookUrl = "n8n_webhook_here"; with your actual n8n webhook URL.
3
Transcript Scraping via n8n
Create an n8n Webhook Use Webhook trigger to receive recordId and Video URL. Call Apify YouTube Scraper Use HTTP Request to: https://api.apify.com/v2/acts/apidojo~youtube-scraper/run-sync-get-dataset-items Pass startUrls, duration, etc., as per your JSON body example. Receive and Parse the Transcript Apify returns XML (or similar structure). Use your Code node to: Extract <text> tags Decode HTML entities Return cleanedText Store Cleaned Transcript Update the original Airtable record with the cleanedText.
4
Optional Preprocessing for RAG
Chunk & Embed Transcript Use Text Split or chunking logic in n8n. For each chunk, embed using OpenAI or similar model. Upload to Pinecone Use HTTP node to push vector embeddings to Pinecone (or Weaviate/Chroma).
5
Transcript Conversation Agent (RAG Chat)
Create a New Webhook or Telegram Trigger Accept a user query and the video ID or reference. Fetch Transcript Chunks from Pinecone Use vector similarity search with the user query. Pass Results to OpenAI Construct a RAG prompt using the retrieved transcript segments. Example system prompt: "Answer the question using the following transcript context." Return Answer Use Telegram or sendMessage node to send the final answer to the user.
6
OPTIONAL: Monitoring and Scaling
Add Status Updates to Airtable Update Status field in Airtable after each step: Scraped Cleaned Uploaded to Pinecone Answered Rate Limiting / Fail Handling Add delays or try-catch blocks for Apify and Pinecone rate limits.
Expected Outcomes
✨
Instant Access to Video Knowledge
Skip the watch time — extract and process full transcripts for on-demand querying.
🎯
RAG-Driven Conversation Flow
Bring your video content to life by letting users ask questions as if chatting with the source.
⏱️
Plug-and-Play for Scale
Integrates with Airtable, Make.com, Pinecone, and Apify — ready for serious automation.
Setup Time
Quick Setup
5 - 10 minutes
Full Configuration
15 - 25 minutes
Last updated
Last Version:
Need Help?
Related Tools
Success Stories
Real results from web3 brands using our automation
NFT Collection
MetaVerse Project
2x Engagement Growth
+15k New Followers
DeFi Protocol
Yield Farming Platform
3x Community Growth
45% TVL Increase
GameFi Project
P2E Gaming Platform
4x Player Acquisition
100k+ Token Holders