✨ $500 AI Visibility Audit — live at Spurlock Studios. Book the audit

Marketing

Youtube Assistant

This n8n workflow functions as a YouTube AI assistant, processing user chat messages to classify their intent (add, summarize, search, or direct response). It intelligently routes requests to either extract, summarize, and store YouTube video data in a Supabase database and…

Built · ~13.5 hours saved per week

YouTube Assistant Workflow Overview This workflow creates a sophisticated, conversational AI assistant for managing and interacting with a personal library of YouTube videos. Users can chat with the assistant to add new videos, ask specific questions about a video they've already added, or perform a semantic search across the entire library to find information discussed within the videos. The automation turns a simple list of YouTube links into an interactive and searchable knowledge base. It automates the entire process of fetching video details, transcribing content, generating AI-powered summaries and tags, and making the content intelligently searchable. Technologies Used - n8n: The core automation platform coordinating all services. - Supabase: Used as the primary data backend for both a relational database (storing video metadata, chat history) and a vector store (for AI-powered semantic search). - Anthropic (Claude 3.5 Haiku): The Large Language Model (LLM) used for all reasoning, classification, summarization, and response generation tasks. - OpenAI: Used specifically for its embedding model to convert video transcripts into vectors for storage and search. - YouTube Data API & Supadata: Services used to fetch video details (title, channel) and transcripts. AI Integration Analysis This workflow is a powerful example of a multi-agent AI system, where different AI components are orchestrated to handle complex tasks. 1. AI as a Router (Intent Classification): The workflow's entry point for any user query is a master AI Agent. This agent's sole purpose is to classify the user's intent (e.g., "add this video," "what was said about Python?", "what can you do?") and output a structured JSON command. This allows the workflow to dynamically route the request to the correct processing path, making the system flexible and reliable. 1. AI for Data Enrichment (Content Generation): When a new video is added, the workflow doesn't just store the link. It uses two dedicated AI agents to process the transcript: - A Summarize Agent creates a concise summary, extracts key points, and identifies actionable tasks.

Tags: Webhook Trigger, Youtube, Google, AI Assistant

Integrations: Supabase, Webhook, AI Agent, Anthropic, Code, HTTP, OpenAI

Related automations

Back to the Automation Library